Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r.com/ locate /pa laeo

Palaeobiogeographic relationships of the Haţeg biota — Between isolationand innovation

David B. Weishampel a,⁎, Zoltán Csiki b, Michael J. Benton c, Dan Grigorescu b, Vlad Codrea d

a Center for Functional Anatomy and Evolution, Johns Hopkins University—School of Medicine, 1830 East Monument St., Baltimore, Maryland 21205, USAb Department of Geology and Geophysics, University of Bucharest, 1 N. Bălcescu Blvd., RO-010041 Bucharest, Romaniac Department of Earth Sciences, University of Bristol, Bristol, BS8 1RJ, UKd Department of Biology and Geology, Babeş-Bolyai University, 1 Kogălniceanu St., RO-400084 Cluj-Napoca, Romania

⁎ Corresponding author. Tel.: +1 410 955 7145; fax:E-mail address: dweisham@jhmi.edu (D.B. Weisham

0031-0182/$ – see front matter © 2010 Elsevier B.V. Aldoi:10.1016/j.palaeo.2010.03.024

a b s t r a c t

a r t i c l e i n f o

Article history:Received 1 May 2009Received in revised form 15 December 2009Accepted 10 March 2010Available online 16 March 2010

Keywords:Haţeg BasinVertebrate faunaLate CretaceousPalaeobiogeographyEvolution

The biogeographic significance of the Late Cretaceous Haţeg fauna is assessed using both faunal andphylogenetic analyses. Although extremely endemic at the species level, the Haţeg fauna is part of a largerEuropean palaeobioprovince compared to roughly contemporary (Campanian–Maastrichtian) terrestrialfaunas elsewhere in Europe. Phylogenetic analyses of five Haţeg taxa, calibrated by biostratigraphicoccurrences provide evidence of long ghost lineages. The geographic distributions of kogaionids,Kallokibotion, Allodaposuchus, and Zalmoxes (together with their European sister taxa) may have arisenfrom vicariant events between western Europe and North America, while the distribution of Telmatosaurus isan example of European endemism of Asiamerican origin.While Haţeg seems to have acted as a dead-end refugium for Kallokibotion and Telmatosaurus, other faunalmembers (and their immediate sister taxa) are not restricted to Transylvania, but known otherwise fromlocalities across southern Europe. In addition, Transylvania may have acted as an evolutionary cradle forkogaionids.Transylvania and the other southern faunas of Europe may represent a distinct division of the LateCretaceous European palaeobioprovince. A boundary between this Tethyan Europe and the more westernand northern cratonic Europe suggests something like the Wallace Line in the Malay Archipelago, in whichtwo distinct faunal provinces with separate histories within a much larger, seemingly uniform geographicregion are separated by a narrow boundary.

+1 410 614 9030.pel).

l rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Understanding the history of any particular extant or extinctecosystem represents one of the major topics of evolutionary studies.It concerns establishing the relationships of that particular assemblagewith its contemporaries fromother areas (i.e., its palaeogeographic ties),as well as those between the same assemblage and those preceding it(i.e., its origin as a complex system of interacting organisms).

In particular, the biogeographic relationships of a given fauna arethe results of the combined effects of changing palaeogeography andorganismal evolution. Continental breakup and assembly, eustaticsea-level changes, tectonic cycles, and climate shifts all influence thegeographic context within which the development of the particularpalaeobiotic unit took place. On the other hand, individual evolution-ary histories of the organisms making up that ecosystem depend ontheir phylogenetic relationships, biological characteristics (such as

dispersal potential, ecological requirements, reproductive strategies),and opportunity (i.e., being in the right place at the right moment).

In the case of the Maastrichtian Haţeg ecosystem, its palaeobio-geographic relationships were probably controlled by its insularsetting within an archipelago-like southern Europe (Nopcsa, 1923a;Weishampel et al., 1991; Dercourt et al., 2000; Benton et al., 2010-thisissue) and the typically continental composition of the fauna (seereviews in Grigorescu, 2005; Benton et al., 2010-this issue),dominated by exclusively terrestrial taxa. The peculiar compositionof the fauna and its palaeobiologic characteristics (Nopcsa, 1923a;Weishampel et al., 1991, 1993, 2003; Csiki and Grigorescu, 2007;Benton et al., 2010-this issue) represent a continuous challenge toour understanding of its palaeobiogeographic relationships.

Like organisms everywhere, the Haţeg vertebrates were theproducts of their individual histories. One aspect of these histories istheir arrival in this region of what is now western Romania. Nopcsa'sapproach to where the dinosaurs (and the remainder of the Haţegfauna) originated was to look solely within Europe. For example, hecompared his Transylvanian hadrosaurid with Iguanodon from therich Early CretaceousWealden faunas of England, Belgium, and France(Nopcsa, 1923a), for the simple reason that both Telmatosaurus and

Table 1Overview of the palaeobiogeographic affinities of the Maastrichtian Haţeg Basin vertebrate assemblage.

Taxon Least-inclusive clade Selected references Speciesa Europeb World Genus Europe World Clade Europe World

Acipenseriform indet. Acipenseriformes Grigorescu et al., 1985 0 x x 0 x x 1 Y; Cz NLepisosteus sp. Lepisosteidae Grigorescu et al., 1999 0 x x 1 N N 1 N NAtractosteus sp. Lepisosteidae unpublished 0 x x 1 N N 1 N NCharacidae indet. Characidae Grigorescu et al., 1985 0 x x 0 x x 1 Y; Pg NHatzegobatrachus grigorescui Anura Venczel and Csiki, 2003 1 Y Y 1 Y Y 1 N NParalatonia transylvanica Discoglossidae Venczel and Csiki, 2003 1 Y Y 1 Y Y 1 N NCf. Eodiscoglossus Discoglossidae Folie and Codrea, 2005 0 x x 1 Y; EK Y 1 N NCf. Paradiscoglossus Discoglossidae Folie and Codrea, 2005 0 x x 1 Y N 1 N NAnura indet. 1 Anura unpublished 0 x x 0 x x 1 N NAlbanerpeton sp. Albanerpetontidae Grigorescu et al., 1999;

Folie and Codrea, 20050 x x 1 N N 1 N N

Kallokibotion bajazidi Kallokibotionidae Nopcsa, 1923b 1 Y Y 1 Y Y 1 N ? YPleurosternon or Polysternon(Pleurodira indet) - large sized

? Bothremydidae Vremir, 2004 0 x x 0 x x 1 N N

Solemydidae indet.(aff. Helocheydra)

? Solemydidae Vremir and Codrea, 2009 0 x x 1 Y; EK N 1 N Y, LJ

Dortokidae n. gen. et sp. Dortokidae Vremir and Codrea, 2009 1 Y Y 1 Y Y 1 N YBicuspidon hatzegiensis Polyglyphanodontinae Folie and Codrea, 2005 1 N Y 1 N Y; EK 1 N Y; EK?Becklesius or ?Paracontogenys Paramacellodidae Folie and Codrea, 2005;

unpublished0 x x 1 Y; EK Y; EK 1 Y; EK Y.; EK

?Beckesius cf. hoffstetteri Paramacellodidae Folie and Codrea, 2005 0 x x 1 Y; EK Y; EK 1 Y; EK Y.; EK?Contogenys Scincidae Smith et al., 2002;

unpublished0 x x 1 Y N 1 N N

?Slavoia Scincomorpha unpublished 0 x x 1 Y N 1 N NSauria indet. 1 Sauria unpublished 0 x x 0 x x 1 N NScincomorpha indet. 1 Scincomorpha Grigorescu et al., 1999 0 x x 0 x x 1 N NScincomorpha indet. 2 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 3 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 4 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 5 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 6 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 7 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 8 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 9 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 10 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 11 Scincomorpha unpublished 0 x x 0 x x 1 N NScincomorpha indet. 12 Scincomorpha unpublished 0 x x 0 x x 1 N NAnguimorpha indet. 1 Anguimorpha Grigorescu et al., 1999 0 x x 0 x x 1 N NAnguimorpha indet. 2 Anguimorpha unpublished 0 x x 0 x x 1 N NAnguimorpha indet. 3(cf. Paraderma sp.)

Anguimorpha unpublished 0 x x 0 x x 1 N N

Anguimorpha indet. 4(cf. ?Paraderma sp.)

Anguimorpha unpublished 0 x x 0 x x 1 N N

Madtsoiidae sp. Madtsoiidae Folie and Codrea, 2005 0 x x 0 x x 1 N NDoratodon sp. Ziphosuchia Grigorescu et al., 1999 0 x x 1 N Y 1 N NAllodaposuchus precedens basal Eusuchia Nopcsa, 1928a; Buscalioni

et al., 20011 Y Y 1 N Y 1 N N

Acynodon n. sp. Alligatoroidea Jianu and Boekschoten,1999; Martin et al., 2006

1 Y Y 1 N Y 1 N N

Musturzabalsuchus sp. Alligatoroidea Jianu and Boekschoten, 1999 0 x x 1 N Y 1 N NZiphosuchia gen et sp. indet. Ziphosuchia Martin et al., 2006 1 Y Y 1 Y Y 1 Y NMesoeucrocodylia gen et sp. nov. Mesoeucrocodylia Martin et al., in prep. 1 Y Y 1 Y Y 1 N NPteranodontidae indet. Pteranodontidae Jianu et al., 1997 0 x x 0 x x 1 Y NHatzegopteryx thambema Azhdarchidae Buffetaut et al., 2002 1 Y Y 1 Y Y 1 N NMagyarosaurus dacus Titanosauria Nopcsa, 1915; von Huene,

19321 Y Y 1 Y Y 1 N N

Titanosauria gen et sp. nov. Titanosauria Csiki et al., in press 1 Y Y 1 Y Y 1 N N“Magyarosaurus” hungaricus Titanosauria Csiki et al., 2007 1 Y Y 1 Y Y 1 N NDromaeosauridae(cf. Saurornitholestes)

Dromaeosauridae Weishampel and Jianu, 1996 0 x x 1 Y N 1 N N

Elopteryx nopcsai Alvarezsauridae Naish and Dyke, 2004;Kessler et al., 2005

1 Y Y 1 Y Y 1 Y; EK N

Euronychodon sp. Troodontidae orDromaeosauridae

Grigorescu et al., 1999;Codrea et al., 2002

0 x x 1 N N 1 N N

Troodontidae indet. Troodontidae Codrea et al., 2002;Smith et al., 2002

0 x x 0 x x 1 Y; EK N

Paronychodon sp. Troodontidae Codrea et al., 2002;Smith et al., 2002

0 x x 1 Y; EK N 1 Y; EK N

Richardoestesia sp. Basal Tetanurae orCoelurosauria

Codrea et al., 2002 0 x x 1 Y; EK N 1 N N

Caenagnathidae indet.(cf. Chirostenotes)

Caenagnathidae Csiki and Grigorescu, 2005 0 x x 0 x x 1 N N

Theropoda 1 Theropoda Unpublished 0 x x 0 x x 1 N NTheropoda 2 (Nălaţ) Theropoda Smith et al., 2002 0 x x 0 x x 1 N NAves Aves Unpublished 0 x x 0 x x 1 N N

420 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Table 1 (continued)

Taxon Least-inclusive clade Selected references Speciesa Europeb World Genus Europe World Clade Europe World

Zalmoxes robustus Rhabdodontidae Weishampel et al., 2003 1 Y Y 1 N Y 1 N YZalmoxes shqiperorum Rhabdodontidae Weishampel et al., 2003 1 Y Y 1 N Y 1 N YTelmatosaurus transsylvanicus Hadrosauridae Weishampel et al., 1993 1 Y Y 1 Y Y 1 N NStruthiosaurus transylvanicus Nodosauridae Nopcsa, 1929 1 Y Y 1 N Y 1 N NBarbatodon transylvanicum Kogaionidae Rădulescu and Samson, 1986;

Csiki et al., 20051 Y Y 1 Y Y 1 Y; Pc Y

Barbatodon sp. (Nălaţ) Kogaionidae Smith et al., 2002 0 x x 1 Y Y 1 Y; Pc YKogaionon ungureanui Kogaionidae Rădulescu and Samson, 1996 1 Y Y 1 Y Y 1 Y; Pc YKogaionon sp. (Nălaţ) Kogaionidae Smith et al., 2002 0 x x 1 Y Y 1 Y; Pc YKogaionon sp. (Toteşti) Kogaionidae Codrea et al., 2002 0 x x 1 Y Y 1 Y; Pc YKogaionon sp. (Pui) Kogaionidae Smith and Codrea, 2003 0 x x 1 Y Y 1 Y; Pc YHainina sp. A Kogaionidae Csiki and Grigorescu, 2000 0 x x 1 Y; Pc Y 1 Y; Pc YKogaionidae gen. et sp. nov. Kogaionidae Csiki and Grigorescu, 2000 1 Y Y 1 Y Y 1 Y; Pc YTheria indet. Theria Csiki and Grigorescu, 2001 0 x x 0 x x 1 N N

a Abbreviations used in taxonomic columns: 0 — indeterminate at the given taxonomic level; 1 — determinate at the given taxonomic level.b Abbreviations used in distribution columns: Y — endemic; N — present in other areas; x — not applicable. Cz — Cenozoic, EK — Early Cretaceous, LJ — Late Jurassic,

Pc — Palaeocene, and Pg — Palaeogene.

421D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Iguanodon, although separated by 50 M.y. and representing phyloge-netically distinct taxa, were European members of Ornithopoda.Likewise, Nopcsa directly compared the other members of theTransylvanian fauna with their European relatives from the EarlyCretaceous as he attempted to understand how his peculiar dinosaursarose. Nopcsa's was a good and logical beginning to our understand-ing of these taxa, but his comparison to other European taxa onlyexplores the possibility of within-continent biogeographic relation-ships, while ignoring those with other continents. In other words, tounderstand where the Haţeg taxa fit in the global biogeographichistory of their clades, it is not enough to consider only Iguanodon,Hypsilophodon, Hylaeosaurus, and Pelorosaurus (and other verte-brates) from Europe, but the net should be spread to include a greatmany more from other continents (as synthesized recently by, e.g.,Kielan-Jaworowska et al., 2004; Weishampel et al., 2004).

Interestingly, one of Nopcsa's lasting contributions to palaeontol-ogy represents an early attempt to understand the evolution of life inthe context of changing palaeogeography (Nopcsa, 1934). Thiscontribution, foreshadowing the ascent of cladistic palaeobiogeogra-phy (Nelson and Platnick, 1981; Wiley, 1988; Grande, 1990), was asurprisinglymodern in approach (see Le Loeuff, 1997), but was largelyoverlooked by subsequent authors. Unfortunately, Nopcsa neverapplied his palaeobiogeographic ideas specifically to the Haţegvertebrate fauna because his life was cut short by suicide.

2. Haţeg palaeobiogeography today

A large amount of new palaeontological and palaeogeographicaldata has been gathered since Nopcsa's original work, both for theHaţeg Basin (see contributions in this issue) and globally. Palaeogeo-graphic reconstructions of the Cretaceous of Europe (e.g., Ziegler,1987; Dercourt et al., 2000; Stampfli and Borel, 2002; Csontos andVörös, 2004; Schmid et al., 2008) and the globe (e.g., Smith et al., 1994;Hay et al., 1999; Scotese, 2004) provide a better understanding of thepalaeogeographic context within which the Haţeg fauna evolved.Discovery of a large number of taxa related to the major groupsrepresented in the Haţeg assemblage on most important continentallandmasses (e.g., Weishampel et al., 2004; Kielan-Jaworowska et al.,2004) has significantly improved our knowledge of the stratigraphicand geographic distribution of the relatives of the Haţeg taxa.Meanwhile, phylogenetic analyses have led to increased accuracy inunderstanding of the phylogenetic relationships of Haţeg taxa.Numerous palaeobiogeographic studies of the Mesozoic (and morespecifically the Cretaceous) continental vertebrates have been pub-lished (e.g., Le Loeuff, 1991, 1997; Russell, 1993; Forster, 1999; Sereno,1999, 2000; Hirayama et al., 2000; Cifelli, 2000; Upchurch et al., 2002;Holtz et al., 2004; Turner, 2004), and the evolution of faunalprovinciality during the Cretaceous is relatively well established.

3. Materials and methods

Currently, the Maastrichtian vertebrate assemblage from Haţegincludes over 70 taxa (Grigorescu, 2005, 2010-this issue; Benton et al.,2010-this issue; see Table 1), ranging from fishes to mammals anddinosaurs. There is a wide variation in taxonomic and especiallyphylogenetic information available for these taxa: while manydinosaurs, turtles, crocodilians and multituberculates are relativelywell understood phylogenetically (Gaffney and Meylan, 1992;Weishampel et al., 1993, 2003; Hirayama et al., 2000; Buscalioniet al., 2001; Pereda-Suberbiola and Galton, 2001; Curry Rogers, 2005;Csiki and Grigorescu, 2006; Delfino et al., 2008), such information is atbest limited for other groups. This places a constraint on the utility ofthese less well understood taxa in palaeobiogeographic analyses:better-known ones are well suited for detailed phylogenetics-basedanalysis, while the others allow only similarity-based faunalcomparisons.

Several palaeobiogeographic analytical techniques have beendeveloped to analyze distribution patterns of fossil organisms. Thesecan be divided roughly into two approaches— faunal similarity-basedand phylogeny-based techniques (Newton, 1990; see Holtz et al.,2004, for a review of these techniques as applied to dinosaurs).Analyses based on faunal similarities employ comparisons of faunallists, without regard for the phylogenetic relationships of the taxarepresented. Some of these techniques rely on simple comparisons ofthe faunas based on overall similarity (e.g., Molnar, 1980; Russell,1993). Other techniques, however, imply a more analytical approach(Le Loeuff, 1991; Holtz et al., 2004), using taxon-occurrence datamatrices to build dendrograms clustering together faunas showingclose biogeographic affinities.

Phylogeny-based approaches develop from the idea that geo-graphic distribution patterns of related taxa should be correlated,among other factors, with their phylogenetic relationships, as thetemporal succession of cladogenetic events must reflect that of thebarrier-forming events that disrupted the unity of once-continuoustaxon distributions. However, barrier formation is one way in whichpalaeobiogeographic units also originate (vicariance), and thus theobserved patterns of cladogenetic events should match to a certainextent the patterns of palaeobiogeographic unit individualization.Events that might blur this direct relationship (regional extinctions,dispersals) can be identified by comparing deduced ages of barrierformation and clade divergence, respectively (e.g., Sereno, 2000). Inthis study, a posteriori optimization analysis of clade origin (e.g.,Weishampel and Jianu, 1997) was used to establish areas of origin forthe most important taxa from the Haţeg Basin for which resolvedcladograms are available.

This palaeobiogeographic analysis concentrates on the Maastrich-tian mammals, turtles, crocodilians, and dinosaurs of the Haţeg Basin,

Fig. 1. A posteriori character optimization A. Assignment of geographic data (Area A) to the terminal taxa (1–4) on a cladogram. The ancestral area of origin for taxa 1 and 2 (Node 5)is most parsimoniously interpreted as Area A, shared with its two descendants. Furher down the tree, taxon 3 and 4 also come from Area A, thus their ancestors (Nodes 6 and 7) areinferred to have come from the same area. This down-pass is called character generalization. B. Optimization double-checks characters back up the tree to resolve any ambiguitiesthat may be left (none in this case). C. Hypothetical example with a clade, half of which (taxa 1–4) come from Area A, while the others (taxa 5–8) from Area B. Generalizing down thetree, the ancestors reconstructed at Nodes 9–11 would also have been found in Area A. The ancestral area at Node 12 could be either Area A or B, so it is registered as “?A/?B.Generalization proceeds to Node 13, where Area B is held in common with Node 12, so the ancestral area at Node 13 is inferred to be Area B. Thereafter down the tree, Area B isinferred as the ancestral area at Nodes 14 and 15. D. Characters are optimized back up the tree. The ancestral area at Nodes 15, 14, and 13 is clearly Area B. Node 13 is then comparedwith Taxon 5 to resolve the ambiguity encountered at Node 12. As Area B is held in common between Node 13 and Taxon 5, the ancestral area at Node 12 is inferred to be Area B. E.With all remaining ancestral areas already identified as A, the shift from Area B to Area A (dispersal) occurred between Node 12 and 11 (E).

422 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

by combining cladistic analysis of these taxa with the globalgeographic occurrences of the clades in which they occur. A posteriorioptimization (Fig. 1) of the geographic occurrences of the Haţeg taxaand close relatives onto their phylogeny requires an explicit, well-supported cladogram. This cladogram is then used to determine themost parsimonious distribution and transformation of the geographicoccurrences (Farris, 1970; Wiley et al., 1991; Brooks and McLennan,1991; Weishampel and Jianu, 1997). Simple and direct a posteriorianalyses can be done by hand, but character optimization on a tree ismore easily accomplished using a numerical cladistic algorithm, suchas PAUP, MacClade, and Winclada.

In this search for the source areas of clades, we rely not only on thepalaeogeographic reconstructions of this part of Europe through theCretaceous, but also those of other relevant landforms. This certainlymeans the palaeogeographic conditions and proximity of parts of Asia,South America, North America, Africa, Antarctica, and Australia.However, inferences about global biogeographic histories can be onlyas good as the fossil record will allow and the biases inherent in thisrecord should be admitted at the start: this is best exemplified by theskewed geographic sampling of dinosaurs throughout the world. Forexample,Asia andEurope contribute about30%of theworld's number ofdinosaur localities for the Early Cretaceous, followed by North Americaat 13% (data fromWeishampel et al., 2004). The remaining continents—SouthAmerica, Africa, and Australia— each contribute nomore than 7%,(not surprisingly, Antarctica has so far contributed nothing). In the LateCretaceous, North America dominates at 36%, followed by Asia at 28%,and South America at 16%; Africa, Australia, and Antarctica togethercomprise a paltry 7%. Even though knowledge of dinosaur distribution

around the world is always on the increase, it is also dominated atpresent by the three continents — North America, Europe, and Asia —

which constitute Laurasia.Consequently, any search for the area of origin of any of the Haţeg

dinosaurs and their immediate clade (as well as those of the otherHaţeg vertebrates) is possibly biased in favour of Laurasia and awayfrom Gondwana, based on the number of fossiliferous locations ratherthan real biogeographic history. Nothing can be done about thisproblem except to extend fieldwork in the southern continents toequalize sampling.

4. Results

4.1. Faunistics

A first approach taken in this study was to compare theMaastrichtian Haţeg vertebrate fauna to other faunas from Europe andworldwide to get a first-hand insight into its possible relationships.

A recent review of the biogeographic relationships of the LateCretaceous continental vertebrates of Europe (Pereda-Suberbiola, 2009)synthesized the large-scale biogeographic affinities of these taxa,including those from the Haţeg Basin. According to this review, mostHaţeg taxa can be identified as showing either Palaeolaurasian (sensuRussell, 1993, representing North America and Asia, eventually Europe:albanerpetontids, discoglossid frogs, paramacellodid and polyglypha-nodontine lizards, basal crocodylians, hadrosaurids, oviraptorosaurs) orEuramerican (covering North America and Europe: alligatoroids,nodosaurids) affinities, while several others (Kallokibotion, dortokids,

423D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

rhabdodontids, kogaionids) are European endemics. Finally, a few othertaxa are considered to point to southern, Gondwanan relationships ofthe fauna, such as madtsoiid snakes and sebecosuchian crocodyliforms(Doratodon); the same claim is, however, dismissed in the case of thetitanosaurians, frequently cited to support Gondwanan affinities of theLate Cretaceous southern European assemblages (e.g., Le Loeuff, 1991).This biogeographic partitioning of the Late Cretaceous Europeancontinental vertebrates falls short, however, in accommodating all thepalaeobiogeographic partitions that can actually be recognized duringthe Cretaceous. For instance, Palaeolaurasia was defined by Russell(1993) as a biogeographic entity covering all northern continents fromthe Early to the latest Cretaceous; however, subsequent studies revealeda different pattern of biogeographic subdivisions during the same timeinterval (see below; Fig. 2). Moreover, lumping together Santonian toMaastrichtian vertebrate assemblages from all over Europe to assignthem to a hypothetical European bioprovince does not recognize thefundamental differences in taxic composition between the differentsectors of the European Late Cretaceous archipelago (e.g., Buffetaut andLe Loeuff, 1991; Le Loeuff and Buffetaut, 1995; Rage, 2002). Conse-quently, the present faunistic analysis focuses specifically on the Haţegfaunal assemblage, without mingling it into a composite Europeanbioprovince.

Fig. 2. Schematic continental palaeobiogeography of the Cretaceous, outlining the main palapresent-day position, drift is not accounted for. A. Late Jurassic: 1—Neopangea, 2— Central Asia — intermittent and selective inter-province exchange route between Asia and Euramerica (eCanudo, 2003); C. Early–Late Cretaceous boundary (Albian–earliest Cenomanian): 5— Asiameexchange routes: b— betweenNorth America and Europe (Weishampel and Jianu, 1997), c— b(Beringia; e.g., Kirkland et al., 1997), e— between Europe and Africa (Canudo et al., 2009 and reD. Late Cretaceous (latest Campanian–Maastrichtian); intermittent and selective inter-provh— betweenNorth America andAsia (Sullivan, 1999; Sereno, 1999, 2000), i— betweenNorth anBuffetaut, 1989; Gheerbrant and Rage, 2006; Pereda-Suberbiola, 2009), k — between Asia and

Geographic resources for our study include the evolving frame-work of biogeographic provinces (Fig. 2) derived from the studies ofBonaparte and Kielan-Jaworowska (1987), Le Loeuff (1991, 1997),Russell (1993), Apesteguía (2002), Upchurch et al. (2002), Krauseet al. (2006), and Upchurch (2008) in conjunction with publishedpalaeogeographic and plate tectonic reconstructions (e.g., Smith et al.,1994; Hay et al., 1999; Dercourt et al., 2000; Golonka, 2004; Scotese,2004; Schmid et al., 2008). According to the proposed reconstruction,most of the Jurassic is characterized by the presence of two majorcontinental bioprovinces: Central Asia and Neopangea, the lastincluding most major continental landmasses except Asia (Russell,1993). Beginningwith the Late Jurassic (Fig. 2A), minor differentiationwithin Neopangea points to the development of a distinct Euramer-ican bioprovince that had intermittent links to the southern,Gondwanan bioprovince through Africa. Whether differentiationwas present during this time within Gondwana is still a matter ofdebate, but was probably already underway (Rauhut and López-Arbarello, 2008 and references cited therein).

During the Early Cretaceous (Fig. 2B), separation of Africa fromEuramerica was completed, and differentiation also began withinEuramerica starting in the Aptian, leading to the individualization of aNorth American and a European bioprovince (Kirkland et al., 1997).

eobioprovinces (M — Madagascar); continental landmasses shown in their approximatea; B. Early Cretaceous (Neocomian–Barremian): 2— Asia, 3— Euramerica, 4— Gondwana;.g., Barrett and Wang, 2007; Barrett et al., 2002; Canudo et al., 2002; Cuenca-Bescós andrica, 6— Greater Gondwana (Apesteguía, 2002); intermittent and selective inter-provinceetween Europe and Asia (Milner and Norman, 1984), d— between North America and Asiaferences therein), f— between Africa and South America (e.g., Buffetaut and Rage, 1993);ince exchange routes: g — between North America and Europe (Martin et al., 2005),d SouthAmerica (Bonaparte, 1984; Gayet et al., 1992), j— betweenEurope andAfrica (e.g.,India–Madagascar (e.g., Prasad and Rage, 1991; Rage, 2003).

424 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Meanwhile, intermittent connections of Europe with Asia allowedlimited faunal exchanges between the two provinces (e.g., Barrett andWang, 2007, and references therein) after a long period of CentralAsian endemism, although the two palaeobiogeographic unitspreserved their identity. Despite the suggestion of the existence ofan Eastern (including India, Madagascar and Australia) and aWesternGondwanan (Africa and South America) province, as hypothesized by,e.g., Le Loeuff (1997), the patchy fossil record of the southerncontinents far from convincingly supports this separation. Instead,Australia at least seems to have been connected to South America assuggested by the presence of closely related taxa during the Aptian–Albian (Smith et al., 2008). Intermittent faunal connections betweenEurope and Africa apparently continued during the Early Cretaceous(Gheerbrant and Rage, 2006), but were probably restricted only topost-Barremian times (Canudo et al., 2009).

The Early–Late Cretaceous boundary is marked by importantpalaeogeographic events (Fig. 2C). The ongoing opening of the NorthAtlantic led to the complete severing of the connections betweenAmerica and Europe (thus the end of the Euramerican province). To thewest, North America became connected to Asia through the rise ofBeringia in the latest Early Cretaceous, leading to a large Asiamericanbioprovince (Kirkland et al., 1997) that lasted into the latest Cretaceous(Sereno, 1999, 2000; Sullivan, 1999; Cifelli, 2000), although regionaldifferences have been recognizedwithin this province (e.g., Godefroit etal., 2003, 2004). Transgression of the epicontinental Western InteriorSeaway resulted in the fragmentation of this province and theindividualization of an Eastern North American biogeographic unit(Appalachia), although faunal support for this unit is still weak(Schwimmer, 1997; Carr et al., 2005). To the south, opening of theSouthAtlantic and IndianOceansdisrupted the continuity of Gondwana.The detailed timing and succession of fragmentation of this palaeogeo-graphic province is controversial (see Sereno et al., 2004; Turner, 2004;Krause et al., 2006), but apparently by the early Late Cretaceous Africabecame isolated from other Gondwanan landmasses, while somedegree of connectionwas probably still present between South Americaand Antarctica–Australia, and India and Madagascar, respectively.

The latest Cretaceous (Fig. 2D) was a time of increased faunalexchanges between different southern and northern landmasses.Such exchangeswere hypothesized to occur between South andNorthAmerica (e.g., Bonaparte, 1984; Bonaparte and Kielan-Jaworowska,1987; Gayet et al., 1992), Europe and Africa (e.g., Buffetaut, 1989;Gheerbrant and Rage, 2006), and Asia and India (Prasad and Rage,1991; Rage, 2003), respectively, but the exact timing, magnitude,direction of the dispersal, and taxa involved are still matters of debate.Dispersal between Europe and North America along a high-latitudecorridor was also suggested by Martin et al. (2005).

Within a global framework, the Haţeg fauna was obviously part ofa larger European bioprovince, based on the presence of a typicalrhabdodontid–titanosaur–basal nodosaurid assemblage (Le Loeuff,1997; Holtz et al., 2004). However, within Europe, it must beemphasized that the Haţeg fauna is markedly endemic and by nomeans a “typical” European Late Cretaceous fauna (whether such anentity can really be defined — see below, Section 5.2).

Although the presence of the Haţeg turtle Kallokibotion bajazid (orclosely related forms) was suggested in western European sites (seeGaffney and Meylan, 1992), none of these occurrences was substan-tiated by more recent studies (Lapparent de Broin and Murelaga,1999). The basal eusuchian Allodaposuchus precedens from Romaniawas also reported to occur in the Late Cretaceous of southern Franceand northern Spain (Buscalioni et al., 1999, 2001); these claims arenot supported by new studies (Martin and Buffetaut, 2005; Delfinoet al., 2008; but see Martin and Delfino, 2010–this issue). Thepresence of the hadrosaurid Telmatosaurus transsylvanicus was citedfrom sites in southern France and northern Spain by Le Loeuff et al.(1993), but these claims were not substantiated more recently(Laurent et al., 1997). Finally, basal euornithopod material from

Haţeg was referred customarily to Rhabdodon priscus (Nopcsa, 1915;Weishampel et al., 1991), but recently was re-interpreted asrepresenting the closely related rhabdodontids Zalmoxes robustusand Zalmoxes shqiperorum (Weishampel et al., 2003).

An overall endemicity analysis was conducted to assess the natureandmagnitude of endemicity of the Haţeg fauna (Fig. 3). From the 70+vertebrate taxa listed from the Haţeg Basin (Csiki, 2005; Grigorescu,2005; Therrien, 2005 and references cited therein; see Table 1), fewerthan two-thirds (42) are sufficiently well known taxonomically to offera meaningful palaeobiogeographic signal. Of these, three taxa are flyinganimals (pterosaurs and birds) whose superior dispersal capabilitiesover potential geographic barriers make them less useful in palaeobio-geographic analysis.

The palaeobiogeographic significance of better-known taxa wasevaluated on their known palaeobiogeographic distribution atEuropean and global scale. Three comparisons were made — atspecies level, at genus level, and at the least-inclusive clade level, inorder to avoid biasing the analysis only towards the cladesrepresented by well-known taxa. Moreover, the dataset was analyzedusing two assumptions, regardless of the taxonomic level involved:first, all taxa present in Haţeg, but missing from the Late Cretaceous ofEurope, and world, respectively, were counted as endemic (conser-vative approach); second, taxa that are represented in Haţeg and not inother areas in the Late Cretaceous, but occur either in older or youngerbeds of Europe, and the world, respectively (called non-coextensiveendemic taxa), were excluded from the list of possible endemisms(relaxed approach). This second assumption identifies the presence ofa non-coextensive endemism as resulting either from regionalextinction (if the respective taxa occur in older beds of other areas),or from the appearance of local evolutionary novelties throughspeciation (if the respective taxa occur in younger beds of otherareas). In the first case, the respective taxa can be considered as late-occurring terminal taxa of declining clades (Lazarus taxa; Jablonski,1986) or local survivors of a regional extinction event (named hereMasada taxa), while in the second case the early-occurring membersof a clade (what might be called Noah taxa) point to possible centresof origin for that respective clade.

This approach provides an idea about the level of endemism in theHaţeg vertebrate fauna, as well as gives hints about its potentialpalaeobiogeographic ties and evolutionary history. It should beunderlined that although, admittedly, this analysis might includecertain flaws (such as those produced by incorrect taxonomicidentification or mistaken stratigraphic and geographic distributionpatterns of the individual clades, resulting from missing data), it canyield a general background for the understanding of the palaeobio-geographic affinities of the fauna.

The analysis revealed a high level of endemism in the Haţeg fauna(Fig. 3). It appears that from the Haţeg taxa known to species level (21taxa, including one aerial taxon— Hatzegopteryx thambema; Buffetautet al., 2002), regardless of the assumption used, none is present in anyother fauna worldwide, except one (Bicuspidon hatzegiensis; Makádi,2006) reported elsewhere in the Late Cretaceous of Europe. Turning togeneric-level taxa (42, including 1 aerial — Hatzegopteryx), using theconservative approach almost three-quarters (30 taxa; 71.4%) of thetaxa are still endemic at a European level, while slightly more (31taxa; 73.8%) are endemic at the global level. Using the relaxedapproach, the level of endemism is, as expected, somewhat lower, butstill considerable: 57.1% (24 taxa) at the level of Europe, and 66.7% (28taxa) globally. In this case, the list of potential endemisms excludespossible Lazarus taxa with earlier occurrences in Europe (Eodisco-glossus, Estes and Sanchíz, 1982), possible local survivors of a regionalextinction (Masada taxa such as Richardoestesia, Paronychodon,known from the Lower Cretaceous of Europe [Rauhut, 2002], butstill present in the Upper Cretaceous of North America [e.g., Sankeyet al., 2002]), as well as Noah taxa (Hainina, known from thePalaeocene of Europe [Vianey-Liaud, 1979]). The shared presence of

Fig. 3. Histogram showing the degree of endemism of the Haţeg vertebrate fauna at species-, genus- and least-inclusive clade level, at European and global scale, respectively.A. Conservative estimate of endemism, considering only the Late Cretaceous presence–absence data of the taxa; B. relaxed estimate of endemism, excluding taxa that are present inHaţeg and absent in Europe/worldwide in the Late Cretaceous, but are represented either in older (Early Cretaceous) or younger (Palaeogene) time slices. Legend: total — totalnumber of taxa from the Haţeg fauna known at the considered hierarchical taxonomic level; endemic Eur — Haţeg taxa endemic at European scale; endemic glob — Haţeg taxaendemic at global scale; indet — indeterminate Haţeg taxa at the respective taxonomic level.

425D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

these taxa between the Maastrichtian Haţeg assemblage and older (oryounger) assemblages in Europe or worldwide suggests palaeobio-geographic connections between Haţeg and other parts of Europeand/or the world before (and after) the Maastrichtian.

Finally, the clade-level analysis still shows a certain amount ofendemism, which is relatively high under the conservative approach(23.9% – 17 taxa – compared to European faunas, 22.5% – 16 taxa –

globally), but drops significantly under the relaxed approach andrelative to Europe (only 2.82% – 2 taxa). This is due to the fact thatmanyhigher-level taxa fromHaţeg occur either in older or younger deposits ofEurope, showing that the origin of these taxa should be sought in theEarly Cretaceous faunas of Europe (Euramerica), or else that certainEuropean Palaeocene taxa might have originated in Haţeg and spreadsubsequently throughout the continent (see e.g., Csiki and Grigorescu,2002). Clade-level endemism compared to areas other than Europe is

still considerably high evenunder relaxed assumption (18.3%–13 taxa),whichmight lend support to the idea that the Haţeg faunawas part of adistinct Late Cretaceous European palaeobioprovince.

To explain this high level of endemism, especially when comparedto late Late Cretaceous (Santonian–Maastrichtian) continental assem-blages from Europe, a high amount of divergent evolution in isolation,producing markedly different local faunas, must be hypothesized.Splitting of the former Euramerican palaeobiogeographic province,followed by extreme fragmentation in southern Europe as a conse-quence of a complex interaction between plate tectonic processes andeustatic sea-level changes (e.g., Tyson and Funnell, 1987; Dercourtet al., 2000; Golonka, 2004) were probably the driving factors behindthis divergent evolution. In order to gain more insight into thetemporal and spatial framework of this process, selected taxa wereanalyzed in their phylogenetic context.

426 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

4.2. Phylogenetic analyses

Implementation of the a posteriori optimization method presentedin the Materials and methods section requires the existence of a well-supported phylogenetic hypothesis that includes the taxon underscrutiny. This prerequisite sets the limits of the following survey, asonly few taxa from the Haţeg vertebrate assemblage have so far beenincluded in phylogenetic analyses. Accordingly, despite the fact that alarge number of endemics can be identified in the Haţeg fauna (seeTable 1), and the potentially important palaeobiogeographic signalthese might offer, taxa such as Paralatonia, Hatzegobatrachus,Elopteryx, or Bicuspidon cannot be discussed within the a posteriorioptimization method. On the other hand, poor skeletal representationof, and taxonomic uncertainty surrounding, taxa such as Acynodon,Doratodon or several small theropods (Paronychodon, Euronychodon,and Richardoestesia) hinder their utility in the same endeavour.Accordingly, only better-known taxa fitted previously into a detailedphylogenetic framework will be considered here; these include thekogaionid multituberculates, kallokibotionid turtles (Kallokibotion),basal eusuchians (Allodaposuchus), and several herbivorous dinosaurs(Zalmoxes, Telmatosaurus, Struthiosaurus, and Magyarosaurus).

4.2.1. MammalsThe mammals of the Haţeg fauna are almost exclusively referable

to the peculiar multituberculate clade Kogaionidae (Rădulescu andSamson, 1996; Csiki and Grigorescu, 2000, 2002, 2006; Codrea et al.,2002; Smith et al., 2002; Kielan-Jaworowska et al., 2004; Csiki et al.,2005); one isolated and fragmentary premolar possibly belongs to anundetermined therian (Csiki and Grigorescu, 2001).

Kogaionids are currently restricted to the Late Cretaceous andPalaeocene of Europe; they occurred exclusively in the Haţeg Basin(and surrounding areas; Codrea et al., 2009a, 2010-this issue) during

Fig. 4. Phylogenetic relationships of derivedMultituberculata (= Cimolodonta) with the posithe ancestors of Arginbaatar from Europe (Euramerica) to Asia (pre-Aptian), 2— isolation of bAptian), 3 — isolation of ancestral kogaionids in Europe, with the fragmentation of EurNorth America to Asia (before the Aptian — the older of the two sistergroups). Abbreviatinumbers: 1 — Early, 2 — Middle, 3 — Late; stages: Alb — Albian, Apt — Aptian, Barr — BarremDanian, Haut—Hauterivian, Kimm— Kimmeridgian, Maa—Maastrichtian, Mon—Montian,San — Santonian, Sin — Sinemurian, Tha — Thanetian, Tith — Tithonian, Val — Valanginian,figures) and chronostratigraphic position of the oldest member is indicated.

the Maastrichtian, but achieved a larger distribution (France, Spain,Belgium, and Romania) during the Palaeocene (Vianey-Liaud, 1979,1986; Gheerbrant et al., 1999; Peláez-Campomanes et al., 2000; seealso Csiki and Grigorescu, 2002). The presence of Hainina was citedfrom Late Cretaceous deposits of eastern North America by Dentonet al. (1996), but more details on this discovery are unavailable and apossible North American occurrence of the kogaionids was not listedby Kielan-Jaworowska et al. (2004). Kogaionids have been included inseveral phylogenetic analyses, but each analysis places them indifferent positions (compare e.g. Kielan-Jaworowska and Hurum,2001, and Rougier et al., 1997, respectively), mainly resulting from theincomplete data sets used by these authors. Only one preliminaryphylogenetic analysis including most members of Kogaionidae and amore complete data set is available (Csiki and Grigorescu, 2006:Fig. 1) which recovered an unresolved Kogaionidae with othercimolodontans (Fig. 4). The distinctiveness of this clade, however, isgenerally accepted (Kielan-Jaworowska and Hurum, 2001; Kielan-Jaworowska et al., 2004).

Almost all Late Cretaceous kogaionids are clustered as basal taxarelative to the more derived Palaeocene representatives of the clade,suggesting a good concordance between their stratigraphic andphylogenetic positions. Moreover, kogaionids are basal to all othercimolodontan multituberculates, but are more derived than Eobaa-taridae, considered to represent one of the most derived clade ofplagiaulacidans, closely related to the ancestry of Cimolodonta(Kielan-Jaworowska and Hurum, 2001; Kielan-Jaworowska et al.,2004). Eobaatarids are known from the Early to late Early Cretaceousdeposits of Europe (England, Spain – Hahn and Hahn, 2002; Kielan-Jaworowska et al., 2004; Badiola et al., 2008), and Asia (China,Mongolia, Japan – Kielan-Jaworowska et al., 1987; Hu and Wang,2002; Kusuhashi, 2008), where they postdate the earliest moments offaunal interchanges between Europe and Asia (Barrett and Wang,

tion of Kogaionidae (modified from Csiki and Grigorescu, 2006). Events: 1— dispersal ofasal cimolodontans in North America, with the fragmentation of Euramerica (before theamerica (before the Aptian), 4 — dispersal of djadochtatherioidean ancestors fromons (used in Figs. 4–10): periods: J — Jurassic, K — Cretaceous, T — Triassic, subscriptian, Ber — Berriasian, Cam — Campanian, Call — Callovian, Cen — Cenomanian, Dan —

Neo— “Neocomian” (undivided early Early Cretaceous), Nor— Norian, Pg— Palaeogene,and Was — Wasatchian. For each terminal taxon, the biogeographic distribution (black

427D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

2007). The oldest occurrence of the group is Valanginian (theeobaatarid Loxaulax, Kielan-Jaworowska et al., 2004). The earliestrecord of cimolodontans more derived than the kogaionids is in theAptian–Albian (oldest occurrence of the paracimexomyid ?Paraci-mexomys crossi, Cifelli, 1997). If kogaionids are indeed the sister groupto other Cimolodonta, the minimal time of origin of Kogaionidae mustbe the Aptian (the age of Paracimexomys).

The place of origin of Kogaionidae is constrained by the distributionof cimolodontans, their immediate sister taxon (eobaatarids), andplagiaulacids (their outgroup after eobaatarids). Most plagiaulacids aredistributed in the Late Jurassic and Early Cretaceous of Europe andNorthAmerica, appearing in the earliest Cretaceous of Asia andnorthern Africa(Morocco), while eobaatarids are restricted to Europe and Asia (Kielan-Jaworowskaet al., 2004).According to this distribution, plagiaulacids aretypical Euramerican taxa with a geographic range extension into Asiaand northern Africa during the earliest Cretaceous; a similar scenario isprobably applicable to Eobaataridae, except they seem to have beenrestricted to the Europeanpart of Euramerica. On the other hand, almostall the Cretaceous derived cimolodontans (with the exception of theAsian endemic Djadochtatherioidea; Kielan-Jaworowska and Hurum,1997) have a distribution restricted to North America during the latestEarly to latest Cretaceous, suggesting they evolved here in isolationbeginning in the Aptian. This date is coincident with the interruption ofthe connections between Europe andNorth America during the Aptian–Albian (Kirkland et al., 1997) and the end of a contiguous Euramericanpalaeobioprovince. Optimization of these distribution patterns onto thecladogram of Csiki and Grigorescu (2006) (Fig. 4) reveals thatCimolodonta had a Euramerican origin sometime during the EarlyCretaceous, followed by a vicariant event leading to the divergencebetween kogaionids (in Europe) and other cimolodontans (in NorthAmerica). According to this evolutionary scenario, Kogaionidae repre-sents an example of European endemism of Euramerican origin.

Be that as it may, the long time gap between the proposed time oforigin of kogaionids during the Early Cretaceous and their lateappearance in the fossil record, in theMaastrichtian, implies anextendedghost lineage (Norell, 1992;Weishampel, 1996)with a duration of about45–50 M.y. This long ghost lineage is similar to that reported for many

Fig. 5. Phylogenetic relationships of basal Cryptodira, with the position of KallokibotionEucryptodira in Asia (before the end of the Middle Jurassic), and 2 — paradoxical appareCretaceous suggested by basal cryptodiran distribution (not represented in Fig. 2).

other taxa from the Haţeg Basin (see Weishampel et al., 1993, 2003),suggesting a common underlying evolutionary history.

4.2.2. Kallokibotion bajazidiWhen described by Nopcsa (1923a,b), K. bajazidiwas considered a

primitive turtle. Apparently restricted to the Transylvanian landmass,Kallokibotion is known only from the Haţeg Basin and the neighbour-ing Transylvanian Basin (Codrea and Dica, 2005; Codrea et al., 2010-this issue). Kallokibotion-like turtles were reported from the Maas-trichtian of European Russia by Averianov and Yarkov (2004), but thisreferral needs more material to be substantiated.

The first phylogenetic analysis of the taxon (Gaffney and Meylan,1992) underscored the primitiveness of the taxon by proposing that itrepresents the sister taxon of Selmacryptodira (all cryptodires morederived than Kayentachelys from the Early Jurassic of North America).Further, more inclusive phylogenetic analyses considering Kallokibotion(Hirayama et al., 2000; Joyce, 2007) recovered this taxon at the base ofthe cryptodiran radiation, in amore basal position thanPleurosternidae.The analysis by Hirayama et al. (2000) found a sister-taxon relationshipbetween Kallokibotion and Tretosternon, a basal cryptodiran consideredby them to include, as synonyms, several other taxa from the earliestEarly to Late Cretaceous (Berriasian–Campanian) of Europe, as well asthe Aptian–Albian and Campanian of North America. Based on Buffetautet al. (1999), Tretosternon-like taxa (Solemys) from the UpperCampanian–Lower Maastrichtian of southern France seem to havebeen present in Europe up to the Maastrichtian.

Optimising the distribution of age and geographic data for basalcryptodirans onto the cladogram of Hirayama et al. (2000) (Fig. 5)suggests that the lineage leading toKallokibotionmust have had its originby theearliest Cretaceous (Berriasian), the ageof theoldest occurrenceofthe Tretosternon-group turtles.Moreover, thisKallokibotion–Tretosternonclade must have had its origin much earlier, since the more derivedXinjiangchelyidae (a subclade of higher crtyptodires) is known from theMiddle–Late Jurassic of Central and Eastern Asia (Peng and Brinkman,1993; Sukhanov, 2001; Matzke et al., 2005). Hirayama et al. (2000)suggested that, based on the presence of basal cryptodirans in the EarlyCretaceous of Australia (Otwayemys), distribution of these basal

(based on Hirayama et al., 2000). Events: 1 — isolation of the ancestors of derivednt close biogeographic relationships between Central Asia and Gondwana during the

Fig. 6. Phylogenetic relationships of basal eusuchians, with the position of Allodaposuchus(simplified after Ősi et al., 2007). Events: 1 — dispersal of Borealosuchus from Europe toNorth America (during or prior to the Santonian), 2 — dispersal of the ancestors ofCrocodylidae from Europe to North America (before the Maastrichtian), based on thehypothesis of the European origin of Alligatoroidea (e.g., Rabi, 2005), 3 — alternativehypothesis, placing the origin of Alligatoroidea+Crocodylidae in North America, after adispersal of their common ancestor fromEurope to North America (before the Santonian),4— isolationof the ancestors of thederivedmesoeucrocodylian Pachycheilosuchus inNorthAmerica, with the fragmentation of Euramerica (after the Barremian).

428 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

cryptodirans must have been nearly global as early as the Middle–LateJurassic.

As noted above, the cladogenetic event separating the Kallokibo-tion lineage from its sister-taxon Tretosternon must be placed no laterthan the earliest Cretaceous (Berriasian, approx. 142 M.y.). Thistiming, together with the geographic distributions of its successiveoutgroups, suggests that Kallokibotion represents the terminus of avery old phylogenetic lineage of Euramerican distribution, a relict inthe Late Cretaceous European fauna (palaeo-endemism); as Hirayamaet al. (2000, p. 192) put it, “…Europe seems have functioned primarilyas a biogeographical “Jurassic Park” or protected area for primitivecryptodires such as pleurosternids, Tretosternon, Kallokibotion, ple-siochelyids, and xinjiangchelyids during the Cretaceous.”

Occurrence of Kallokibotion late in the stratigraphic record (Maas-trichtian, approx. 70–68 M.y.) implies the existence of an extendedghost lineage of about 70 M.y., apparently much longer than in the caseof any other taxon from the Haţeg Basin. Despite such a long hiddenhistory,Kallokibotion seems to have undergone onlyminor evolutionarychange, and its restricted geographic range (Transylvanian landmassand perhaps southwestern Russia) suggests that this might have beendrivenbyvicarianceprocesses after the separation fromtheTretosternonlineage that was distributed in western Europe and North America.

On the other hand, the analyses of Joyce (2007) and Danilov andParham (2008) found Kallokibotion to lie outside crown-groupTestudines, as a relatively derived stem testudinate. Considering thisalternative, more basal position of Kallokibotion does not alter thepalaeobiogeographic scenario outlined above. Instead, shifting Kallo-kibotion into a more basal position within Testudinata only underlinesthe antiquity of the evolutionary lineage leading to it, since its sister-taxon (Paracryptodira+crown-group testudines) is already knownfrom the Middle Jurassic (Danilov and Parham, 2008). Accordingly,the ghost lineage leading to Kallokibotion extends from the MiddleJurassic (probably Bathonian–Callovian; see Weishampel et al., 2004and references therein) to the Maastrichtian, spanning an even moreimpressive time interval of about 90–95 M.y. According to theseauthors, the replacement of stem testudines with stem cryptodiresand even crown-group Cryptodira in Europe was already under wayby the Late Jurassic, and thus survival of a basal testudinate lineageinto the latest Cretaceous is even more remarkable.

4.2.3. A. precedensNopcsa (1928a) proposed the name A. precedens for crocodilian

remains he described from the Maastrichtian of the Haţeg Basin(Nopcsa, 1915), invoking the possibility that in the future a specificdifference should be established with the crocodilian remains fromFuveau, southern France, known at that time as Crocodylus affuvelensis(see Martin and Buffetaut, 2008). Specimens referred to Allodapo-suchus were described from the Upper Cretaceous of southern Franceand northern Spain (Buscalioni et al., 1999, 2001), but theseidentifications were questioned (Martin and Buffetaut, 2005; Delfinoet al., 2008). Currently, the taxon is known only from Transylvania,although other species of Allodaposuchus might have been present inwestern Europe (Delfino et al., 2008; Martin, in press).

The first phylogenetic analysis to include Allodaposuchus (Busca-lioni et al., 2001) suggested that it occupies a basal position withinEusuchia, representing the sister taxon of crown-group Crocodylia.More recent analyses, based on larger character-taxon matrices(Salisbury et al., 2006; Ősi et al., 2007) and more complete referredmaterial of Allodaposuchus (Delfino et al., 2008) group Allodaposuchuswith Hylaeochampsa from the Early Cretaceous (Barremian) of Europe(Clark and Norell, 1992), in an unresolved tritomy with Hylaeo-champsidae and the crown-group, or even into moving it into aslightly more basal position as the sister group of (Hylaeochampsa+crown-group). Despite this uncertainty regarding the exact phyloge-netic position of Allodaposuchus, its phylogenetic proximity toHylaeochampsidae at the base of Eusuchia seems well established.

Based on this phylogenetic position (Fig. 6), the age and place oforigin of the Allodaposuchus lineage can be estimated by optimising thestratigraphic and geographic distribution of eusuchians and their closestrelatives onto the cladogram of Delfino et al. (2008). A sister-grouprelationship between Hylaeochampsa and Allodaposuchus suggests thatdivergence between the two lineages occurred at least by the Barremian(at approx. 130 M.y.). During the Mesozoic, crown-group crocodyliansare restricted to the Late Cretaceous, and the dominantly NorthAmerican distribution of the oldest members of the clade (e.g. Brochu,1999; Salisbury et al., 2006) was used to suggest a North Americanorigin. However, the presence of basal alligatoroids in the Santonian–Campanian of Europe challenges this view (Rabi, 2005; Martin, 2007;Martin and Buffetaut, 2008) and, together with the European distribu-tion of the hylaeochampsid sister group (Clark and Norell, 1992; Ősi etal., 2007) to the crown-group, suggests a possible European origin of theancestral crown-group crocodylians. Most of the successive advancedneosuchianoutgroups to theHylaeochampsa–Allodaposuchus–Crocodyliagroup (e.g., Bernissartia, Goniopholis, and Theriosuchus) have a Euramer-icandistribution during the Late Jurassic–Early Cretaceous, and suggest aEuramerican origin of the derived neosuchians. Recent discoveries ofGondwanan late Early Cretaceous non-crocodylian eusuchians (Salis-bury et al., 2003, 2006) were cited as arguments for a possible southernlocation of the Neosuchia–Eusuchia transition. However, their latestratigraphic occurrence, compared to that of the next outgroup taxa ofEusuchia (Late Jurassic) predating the definitive fragmentation ofNeopangea, suggests that the Gondwanan presence of mesoeucrocody-lians of neosuchian-basal eusuchian grade can be probably explained byvicariant processes fragmenting an already wide Neopangean distribu-tion. Choosing between these palaeobiogeographic scenarios might restultimately on the relative phylogenetic positions of Susisuchus andIsisfordia, recovered as a polytomy with Goniopholis and Bernissartia atthe base of Eusuchia by Delfino et al. (2008), but as a clade more basalthan Hylaeochampsa by Martin and Buffetaut (2008). New discoveries,especially from the more poorly sampled southern continents mightshedmore light on the exact timing and location of the events leading to

429D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

the origin of Eusuchia, but it is unlikely that these will challenge theideas of a Neopangean origin of the Neosuchia, nor of a Euramericanorigin of the advanced Eusuchia during the Early Cretaceous, before theBarremian.

Allodaposuchus, as a result, is recognized here as an endemic taxon,descending from a Euramerican common ancestor shared withHylaeochampsa. Evolution of Hylaeochampsidae continued in theEuropean area after the post-Barremian fragmentation of Euramerica,leading to Iharkutosuchus in the Santonian of Hungary (Ősi et al., 2007),while another closely related lineage peaked in Allodaposuchus from theMaastrichtian of Romania The late stratigraphic occurrence of Alloda-posuchus suggests the presence of a long ghost lineage of about 55–60 M.y., similar to other vertebrates known from theHaţeg Basin, thus along period of hidden history. (An alternative evolutionary scenariowould be required if Allodaposuchus shifts to a more derived positionwithin basal alligatoroids, according to Martin (in press). However, thisscenariowill not be further explored here, pending the studies ofMartinon a possible new species of Allodaposuchus in southern France).

4.2.4. Zalmoxes speciesUntil recently, nearly all basal ornithopods from the Late Cretaceous

of Europe were assigned to Rhabdodon (Matheron, 1869; Nopcsa, 1902,1915, 1928b;Weishampel et al., 2003). Ofmodest size and robust build,these dinosaurs are known from present-day France, Spain, Austria,Romania, and Hungary. In 2003, Weishampel et al. revised the basalornithopod material from Haţeg referred previously to Rhabdodon andanalyzed its phylogenetic significance, noting that the Transylvanianmaterial was different from that elsewhere in Europe. Called Zalmoxes,two species were recognized: Z. robustus and Z. shqiperorum. Both sharethe same stratigraphic distribution and geographic provenance — theLate Cretaceous of present-day western Romania; subsequently, thegenuswas also reported fromthe early Campanianof Austria (Sachs andHornung, 2006). Its closest relative, R. priscus from southern France andnorthern Spain, appears roughly contemporaneous with the twoZalmoxes species from Transylvania (Weishampel et al., 2003). Fromthis, it can be concluded that there was a negligible ghost lineage in thisclade of rhabdodontids (which did not include the new material fromHungary) and that this clade originated in Europe (Fig. 7).

Fig. 7. Phylogenetic relationships of basal Ornithopoda, with the position of Zalmoxesand Rhabdodontidae (based on Weishampel et al., 2003; Butler et al., 2006, 2008).Events: 1 — dispersal of the ancestors of Late Cretaceous American basal ornithopods(Thescelosaurus, Parksosaurus, and Gasparinisaurus) from Europe to the New World(during or prior to the Kimmeridgian), 2— isolation of the ancestors of Tenontosaurus inNorth America, with the fragmentation of Euramerica (after the Barremian).

The sister-group of Rhabdodontidae, Iguanodontia, is a clade thatdates back to the Late Jurassic (the earliest ages of Camptosaurus andDryosaurus; probably near the boundary between the Kimmeridgianand Tithonian, approximately 151 mya) of western North America(Norman, 2004). Beyond that, several reasonably well-known taxaformerly thought to have been basal ornithopods (Agilisaurus,Hexinlusaurus, and others) have been shifted to more basal positionswithin Ornithischia (Butler et al., 2008). With these removed, siblingrelationships are unresolved more basally than Rhabdodontidae+Iguanodontia, with Thescelosaurus, Parksosaurus, Gasparinisaura, andthe latter clade forming a polytomy. Relationship with Hypsilophodonfollows thereafter. Parksosaurus and Thescelosaurus have a NorthAmerican distribution, while Gasparinisaura is South American andHypsilophodon is European. Taken together, these dinosaurs indicatethat the most parsimonious of source areas of the entire clade isEuropean (Euramerican, taking into account the age of the basalHypsilophodon), with dispersal of Gasparinisaura, Thescelosaurus,Parksosaurus, and Iguanodontia to the New World. Zalmoxes andRhabdodon are part of the original European distribution (Fig. 7).

4.2.5. Telmatosaurus transsylvanicusNopcsa originally described what he considered hadrosaurid

material from the Haţeg Basin in 1900 at the age of 22. Properlynamed T. transsylvanicus in 1903, Nopcsa considered this species to bepositioned at the base of Hadrosauridae. In addition to Haţeg,specimens referred to Telmatosaurus have also been claimed in Franceand Spain (Lapparent, 1947; Le Loeuff et al., 1993), but they have notstood up to further investigation (Laurent et al., 1997). CurrentlyTelmatosaurus is known only from Transylvania (Dalla Vecchia, 2006).

The first cladistic analysis to include Telmatosaurus, (Weishampelet al., 1993) confirmed Nopcsa's (1900) suggestion that it was the basalmember of Hadrosauridae; that is, the sister-group of the cladeconsisting of lambeosaurines and hadrosaurines. This position hasbeen confirmed over the ensuing years through the discovery of manynew hadrosaurids and the application of cladistic methods with largercharacter-taxon matrixes to understand the general shape of thecladogram of the group (Weishampel et al., 1993; Godefroit et al.,1998; Norman, 2002, 2004; You et al. 2003a,b; Horner et al., 2004;Fig. 8). More recently, Forster (1997) argued that Hadrosauridae shouldconsist solely of Lambeosaurinae and Hadrosaurinae, which placesTelmatosaurus outside this clade. This alternative position of Telmato-saurus, just outside Hadrosauridae sensu stricto, was re-iteratedrecently by Sues and Averianov (2009). In both cases, the relationshipof Telmatosaurus, lambeosaurines, and hadrosaurines is maintained andthe issue becomes semantic (see Horner et al., 2004). Moreover, whenanalyzing the cladogram that depicts the preferred phylogenetichypothesis of Sues and Averianov (2009), only details such as time oforigin and dispersal of the Telmatosaurus lineage to Europe differ,compared to those derived from our preferred phylogenetic hypothesis.

Based on this phylogenetic position, with the closest relatives ofTelmatosaurus being the more commonly known clade of Hadrosaur-inae+Lambeosaurinae — called Euhadrosauria by Weishampel et al.(1993) — and successively more distant Asian and North Americanoutgroups, the Telmatosaurus lineage must be as old as or older thanthe late Albian (the age of “Trachodon” cantabrigiensis, an indetermi-nate euhadrosaurian from England; Lydekker, 1888).

These relationships imply a ghost lineage for Telmatosaurus of30 M.y.; the alternative hypothesis of Sues and Averianov (2009)suggests a somewhat shorter ghost lineage, of about 18 M.y.

Beyond the timing of these cladal origins, the palaeobiogeographyof Hadrosauridae requires geographic locations of taxa farther downthe cladogram. Unfortunately, the sister group of Euhadrosauria+Telmatosaurus (Hadrosauridae sensu Weishampel et al., 1993) as wellas successive outgroups are not altogether certain. A few studies haveindicated that forms like Eolambia caroljonesa and Protohadros byrdi(Kirkland, 1998; Head, 1998; You et al., 2003a), both from the earliest

Fig. 8. Phylogenetic relationships of ankylopollexian ornithopods, with the position of Telmatosaurus (based onWeishampel et al., 1993; Horner et al., 2004; Butler et al., 2006). Events:1— isolation of the hadrosauroids in Asiamerica (after the Barremian), 2— dispersal ofOuranosaurus from Europe to Africa (pre-Aptian; but see Canudo et al., 2009); 3— dispersal of basalhadrosaurines from Asia to North America (prior to the Campanian); 4 — dispersal of derived lambeosaurines from Asia to North America (in several episodes after the Santonian, withoccasional reversal of the dispersal direction; see also Godefroit et al., 2003); 5 — dispersal of Telmatosaurus from Asiamerica to Europe (probably before the Albian; see text).

430 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Late Cretaceous (approximately 95 mya) of North America, are thenext successive close relatives of Hadrosauridae. More distantrelationships beyond this point are with a variety of Asian andNorth American taxa. Other, more recent work places one or severalAsian taxa (among them, Bactrosaurus, Probactrosaurus, and Nanyan-gosaurus; Godefroit et al., 1998; Xu et al., 2000; Norman, 2002) assister taxa of Hadrosauridae.

When locations of these taxa are optimized on the cladogramsummarizing current understanding of higher ankylopollexian relation-ships (Fig. 8), the most parsimonious source area of the entire clade isEuramerica. The evolution of the clade of Probactrosaurus and highertaxa involves dispersal to Asia, with a second dispersal of theTelmatosaurus lineage to Europe. Finally, the Euhadrosauria/Hadrosaur-idae clade has an Asiamerican source. The notion of a North Americanorigin for Hadrosauridae has a long history (e.g., Lull andWright, 1942;Ostrom, 1961), but when Asian taxa are intermixed with those fromNorth America as successive sister taxa to Hadrosauridae (as indicatedabove), then the source area becomes Asiamerica.

The European location of Telmatosaurus, combined with itsrelationship among ankylopollexian taxa, stands out against theotherwise Asian/Asiamerican distribution of closely related taxa. Inlieu of the discovery of new taxa, Telmatosaurus is best interpreted asthe terminus of a single European migration.

4.2.6. Other dinosaursLess is known about the relationships of the nodosaurid Struthio-

saurus transylvanicus, a member of a clade of three described speciesdistributed from Transylvania to Austria and the Ibero-Armoricanlandmass (see Garcia and Pereda-Suberbiola, 2003, Pereda-Suberbiolaand Galton, 2009). The phylogenetic position of Struthiosaurus issomewhat uncertain; although usually considered a nodosaurid (e.g.,Pereda-Suberbiola and Galton, 2001), this was considered not wellsupported by Vickaryous et al. (2004). It was recently included in twophylogenetic analyses (Ősi, 2005;Ősi andMakádi, 2009) that resolvedits position at the base of Nodosauridae, followed by Hungarosaurus

from the Santonian of Hungary, and, further up the tree, by a clade oflate Early to Late Cretaceous North American taxa (Silvisaurus,Sauropelta, Pawpawsaurus, Edmontonia, and Panoplosaurus); unfortu-nately, its relationships to Cedarpelta, considered the oldest (Berria-sian–Hauterivian of North America) and basalmost nodosaurid byVickaryous et al. (2004) is unknown at present. Phylogeneticrelationships of S. transylvanicus within the genus have not yet beeninvestigated.

Based on available phylogenetic data (Fig. 9), the origin of theStruthiosaurus lineage extends into pre-early Aptian times (ghostlineage duration about 44 M.y.), but could go back as far as the earliestCretaceous, depending on its relationships with Cedarpelta (estimatedghost lineage duration up to 75 M.y.). However, with its sister speciesranging only as far as the Early Campanian (Garcia and Pereda-Suberbiola, 2003; Pereda-Suberbiola and Galton, 2001), the origin of S.transylvanicus itself extends back only to the Santonian–Campanianboundary; this suggests the presence of land connections between thedifferent southern European landmasses during the second half of theLate Cretaceous. As for the origin of the Struthiosaurus lineage, aposteriori optimization using the tree topology found by Ősi andMakádi (2009) suggests that it (together with that of Hungarosaurus)can be traced to Euramerica, with their successively more derivedrelatives living in the Early Cretaceous of North America (Vickaryouset al., 2004). Thus, it seems reasonable that the common ancestor ofStruthiosaurus and its closest relatives was Euramerican and dates toat least the late Aptian. The identification of a clade of derived lateEarly to Late Cretaceous North American nodosaurids suggests thatisolation following the severing of land connections between theeastern and western parts of Euramerica (i.e., Europe and NorthAmerica, respectively) might have been the driving factor for theendemic development of the Struthiosaurus lineage in Europe.Subsequently, eustasy-driven fragmentation of the different southernEuropean landmasses might have produced the split of this lineageinto the three known species inhabiting different islands. However,poor skeletal representation of the Transylvanian Struthiosaurus,

Fig. 10. Phylogenetic relationships of Ankylosauria, with the position of Struthiosaurus (based on Vickaryous et al., 2004; Ősi, 2005; Ősi and Makádi, 2009). Events: 1— dispersal and/orisolation of the derived nodosaurids in North America after the Barremian, following the severing of the Euramerican land connections, 2 — dispersal of the ancestors of higherAnkylosauridae from Euramerica to Asia (before the Aptian), 3— dispersal of the ancestors ofMinmi from Euramerica to Australia (before the Aptian), and 4— reintroduction of themostderived Ankylosaurinae from Asia to North America (before the Campanian).

Fig. 9. Phylogenetic relationships of Macronaria, with the position of Magyarosaurus (based on Curry Rogers, 2005; Upchurch et al., 2004). Events: 1 — dispersal of the ancestors ofIsisaurus from Laurasia to India (prior to the Maastrichtian); 2 — dispersal of the ancestors of ‘Rapetosaurus clade’+Saltasauridae from Laurasia to Gondwana (Africa) before theAptian, 3 — dispersal and isolation of the ancestors of Magyarosaurus before the Aptian, 4 — reintroduction of the ancestors of Nemegtosaurus into Asia before the Maastrichtian.

431D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

432 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

together with the uncertainties regarding the exact phylogeneticposition of the genus (compare Ősi andMakádi, 2009 with Vickaryouset al., 2004 and Carpenter, 2001), suggests that this palaeobiogeo-graphic scenario must be treated with caution.

Unfortunately, far less is known about the relationships of theTransylvanian titanosaurs (Magyarosaurus and yet to be namedforms) — where they might have originated and when they diverged —

but efforts to obtain this information is now underway (e.g., Csiki et al.,2007, in press). Only one phylogenetic analysis including Magyarosaurushas been published so far (Curry Rogers, 2005). According to this study,Magyarosaurus clusters as a member of an informal ‘Rapetosaurus clade’(Fig. 10) alongside Rapetosaurus from the ?Campanian or ?Maastrichtianof Madagascar, Trigonosaurus from the Maastrichtian of Brazil,Malawisaurus from the Aptian of Malawi and Nemegtosaurus fromthe Maastrichtian of Mongolia (Upchurch et al., 2004). The sistergroup of the ‘Rapetosaurus clade’ is the exclusively South AmericanLate Cretaceous (Santonian–Maastrichtian) saltasaurine clade(Saltasaurus, Neuquensaurus, and Rocasaurus). Based on this topology,Magyarosaurus represents a late-surviving offshoot of an evolutionarylineage dating back to at least the Aptian, yielding a ghost lineageduration of about 50 M.y. Moreover, optimization of the biogeographicarea of origin suggests a southern, Gondwanan (western Gondwanan)origin for the ‘Rapetosaurus clade’, with introduction of the evolutionarylineage leading to Magyarosaurus from western Gondwana (probablyAfrica) into Europe some time during the Aptian or later. Interestingly,the earliest probable timing of this faunal exchange coincides closelywith the opening of a dispersal route between Africa and Europe in theBarremian-Aptian, as hypothesized recently by Canudo et al. (2009).

Although the available data provide the basis for a posteriorioptimization to determine the time and place of origin for Magyar-osaurus, several lines of evidence suggest extreme caution must betaken in this respect. First, Magyarosaurus was coded in Curry Rogers(2009) on the assumption that all titanosaur remains from the Haţegbelong to one taxon (e.g., Le Loeuff, 1993). However, a recent revisionof the available material suggests the presence of several taxa (e.g.,Csiki et al., 2007, in press), which casts doubt on the codingssupporting the phylogenetic position ofMagyarosaurus. Moreover, theposition of Magyarosaurus within its clade is unstable and poorlysupported (see Curry Rogers, 2005, for details); consequently anypalaeobiogeographic hypothesis based on this analysis rests on ratherweak grounds. These phylogenetic uncertainties surroundingMagyar-osaurus, and other taxa not yet included in cladistic analyses, suggestcaution until a full revision of the Haţeg titanosaurs is published.

5. Discussion

From the foregoing, the biogeographic affinities of the Haţeg faunapoint to a panoply of sources. First, in terms of faunistics, Haţeg isstrongly endemic with respect to other European faunas. Second,phylogenetic analyses indicate a mixture of dispersal and vicariantexplanations. Dispersal fromAsiamerica accounts for the European (i.e.,Transylvanian) distributionof Telmatosaurus. On the otherhand, a broadEuropean/Euramerican distribution argues for an early isolation, beforethe end of the Early Cretaceous, of core Late Cretaceous European taxa(e.g., multituberculate mammals, basal cryptodirans, hylaeochampsid-grade eusuchian crocodylians, basal ornithopods), with subsequentvicariant events involving Transylvanian taxa.

The above discussions suggest a complex palaeobiogeographichistory of the Haţeg assemblage. The wider implications of thesediversified palaeobiogeographic affinities, as outlined, will be dis-cussed in the following.

5.1. The Haţeg Island — a refugium?

Insularity is often thought to represent an evolutionary dead-end,islands acting as refugia for the terminal members of a clade. The last

surviving mammoths (Mammuthus primigenius) were reported tohave inhabited small islands in the Arctic Ocean (Vartanyan et al.,1993, 1995, 2008; Guthrie, 2004). New Zealand and its neighbouringsmaller islands represent sanctuaries within which endemic repre-sentatives of several primitive clades (e.g., Leiopelmatidae, Spheno-dontia) survive (Mitchell et al., 2006; Worthy et al., 2006; Jones et al.,2009). Finally, a peculiar vertebrate assemblage described from theAlbian Tlayua Formation, Mexico, including several basal forms(squamates, scincoids), was interpreted as an insular fauna (Reynoso,2000).

Relatively early, Nopcsa (1915, 1923a) interpreted the Haţeg faunain similar terms, as an assemblage composed of primitive taxasurviving within an island sanctuary, shielded from competition frommore derived contemporaries by occupying an isolated region.

In the cases ofmost of the taxadiscussed above, suchanevolutionaryhistory is certainly conceivable, especially in the case of those taxa forwhich a long ghost lineage can be reconstructed, and whoseunexpectedly late stratigraphic occurrences contrast with their strik-ingly primitive morphology. Kallokibotion or Telmatosaurus certainly fitthe case of survivorswithin an isolated island refuge. But even these twotaxa imply different interpretations: while for Kallokibotion the Haţegarea acted as a palaeorefugium, an area circumscribing a fragment of aonce-existing wider distribution (Nekola, 1999), for Telmatosaurus itrepresents a neorefugium, a sanctuary area formed after the geographicrange expansion of the basal hadrosaurids into Europe (Nekola, 1999).

It should be emphasized, however, that the palaeobiogeographicsignificances of other members of the Haţeg fauna do not comply withsuch a simplistic refugium model. Although rhabdodontids andAllodaposuchus, together with Struthiosaurus and several other lowervertebrate taxa (i.e., Bicuspidon) represent isolated and late-survivingdescendants of very old phylogenetic lineages, these taxa are by nomeans restricted to the Transylvanian area. All show awider, southernEuropean distribution, encountered on other landmasses as well(Weishampel et al., 2004; Makádi, 2006; Martin, in press). Here, theyare represented eventually by different species, or closely relatedsister taxa, suggesting local evolution and speciation on the differentlandmasses. In some cases, the Haţeg taxa and their relatives fromother parts of southern Europe are more or less contemporary,removing the Transylvanian taxa as the latest survivors of their clades.Accordingly, these taxa do not fit the model of insular relicts, butrepresent instead members of an endemic European faunal radiationimplying clades isolated here after the Early Cretaceous. Moreover,presence of closely related, even sister-species taxa, across differentparts of southern Europe calls for the presence of dispersal routesbetween these isolated landmasses. The time of opening of thesedispersal routes, as well as their exact palaeogeographic position,nature, and selectivity, are poorly constrained and should representan important direction for future studies of Late Cretaceous Europeanpalaeobiogeography.

Several other, lesswell-known clades fromHaţeg are represented bytaxa with a widespread distribution during the Late Cretaceous (seereview in Pereda-Suberbiola, 2009); these include lepisosteids, dis-coglossids, albanerpetontids, basal alligatoroids, dromaeosaurids andother theropods with uncertain affinities (Richardoestesia and Parony-chodon). Whether these represent descendants of older European taxaor immigrants from other (mainly Asiamerican) landmasses, remainsundetermined until more diagnostic material is discovered that willallow a thorough comparison with their relatives from other palaeo-biogeographic provinces. Even theropodswhose affinities seem to pointto Asiamerican connections during the Late Cretaceous (i.e., alvarez-saurids and caenagnathids; Osmólska et al., 2004; Padian, 2004;Longrich and Currie, 2009) have been interpreted previously as possibledescendants of Early Cretaceous members of these clades present inEurope/Euramerica (Csiki and Grigorescu, 2005; Kessler et al., 2005).

Finally, the “Haţeg Island” can be interpreted as an evolutionarycradle (see Bellemain and Ricklefs, 2008 for modern examples) for at

433D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

least kogaionid multituberculates. According to their known strati-graphic and geographic distribution, kogaionids were restricted to theTransylvanian area during the Late Cretaceous, but spread acrosssouthern Europe during the Palaeocene (Csiki and Grigorescu, 2002;Kielan-Jaworowska et al., 2004; Csiki et al., 2005 and references citedtherein). Based on this distribution pattern, as well on the apparentlybasal position of most Transylvanian kogaionids compared to thePalaeocene representatives of this clade, it is tempting to hypothesizethat the Transylvanian landmass, isolated for most of the LateCretaceous from other emerged areas, represented the place ofdiversification, if not the origin, of Kogaionidae (Csiki and Grigorescu,2002).

5.2. Haţeg Island and the Late Cretaceous European palaeobioprovince

As already noted in the faunal comparisons (4.1.), the overallcomposition of the Haţeg fauna is reminiscent of those from otheremergent parts of southern Europe. However, although Late Creta-ceous Europe is usually considered a coherent bioprovince in mostmodels (Le Loeuff, 1997; Holtz et al., 2004), its unity is far from wellsupported (e.g., Upchurch et al., 2002). The faunal differences acrossthe Late Cretaceous of Europe (e.g., Buffetaut and Le Loeuff, 1991; LeLoeuff and Buffetaut, 1995; Rage, 2002) can be ascribed both to theextreme geographical fragmentation of the region (Dercourt et al.,2000), and possibly to the differential palaeobiogeographic evolutionof the segments of this bioprovince.

Geographically, it seems difficult to treat the different LateCretaceous continental faunal assemblages of Europe as parts of aunique and homogenous bioprovince; this is contradicted by thepalaeogeographic-palaeotectonic setting of Late Cretaceous Europe.The advanced geographical fragmentation of southern Europe aroseboth from the presence of epicontinental seaways transgressed overthe stable European craton, and to the different deep-sea basinsconnected to the Neo-Tethys and Alpine Tethys oceans (e.g., Dercourtet al., 2000; Stampfli and Borel, 2002; Golonka, 2004; Schmid et al.,2008). An analogue probably can be found in the modern MalayArchipelago extending between mainland Asia and Australia.

A short survey of the most outstanding faunal differences revealsinteresting distribution patterns across Europe. Despite the relativelyhigh diversity of the Haţeg fauna, it is noteworthy that it lacksmembers of several clades represented in western European verte-brate faunas. For example, hadrosaurids appear to have been quitecommon in latest Cretaceous assemblages from Ibero-Armorica, withdiverse basal lambeosaurines as a newly recognized hallmark feature(Pereda-Suberbiola et al., 2009; Prieto-Marquez and Wagner, 2009);these taxa are missing from hadrosaurid-bearing assemblages fromsoutheastern Europe (Romania, Italy). Abelisauroid theropods weredescribed from areas belonging to the partly submerged southernmargin of the European mainland (France, the Netherlands; seeCarrano and Sampson, 2008 and references therein), but not frommore eastern areas, evolving as insular continental blocks within thenorthern part of the Tethys (Austria, Hungary, Italy, Slovenia,Romania). Rare records of European Late Cretaceous ceratopsians areknown from north-western European sites (Godefroit and Lambert,2007; Lindgren et al., 2007), but these are absent frommore southernand eastern (including Haţeg) assemblages. The eutherian zhelestidsare well represented in the Ibero-Armorica realm (see review byKielan-Jaworowska et al., 2004), while raremarsupials were describedfromnorthwestern Europe (theNetherlands;Martin et al., 2005); bothof these groups are absent from the eastern European faunas, includingHaţeg. Besides these differences in faunal composition, each landmassharboured an array of endemic taxa, differentiated at least at specific, ifnot generic level (see Pereda-Suberbiola, 2009). Taken together, thesefaunal differences require a two-step hypothesis to explain: minordifferences (clades represented by different species or genera ondifferent European areas) most probably arose from differentiation

from a common core assemblage, or by accidental dispersal ofindividual taxa from other palaeobioprovinces, whilemajor differences(entire clades present in certain areas and absent from others)represent more substantial and more recent palaeobiogeographicties with neighbouring palaeobioprovinces.

According to these alternative explanations, the following predic-tions can be advanced:

1. Minor differences will occur randomly between the different areas,because they result from local extinctions or speciations experi-enced by the different clades, or from random dispersal acrossexisting barriers.

2. Major differences will occur non-randomly, as members ofdifferent clades appear associated recurrently in certain areas,while are missing, again recurrently, in others.

3. Clades occurring recurrently together share a common palaeobio-geographic affinity, usually pointing outside, towards a neighbour-ing palaeobioprovince. The identification of a dispersal routebetween the two areas, supported by independent (geological–geophysical) evidence, is to be expected.

Surveying the composition of the different Late Cretaceous localfaunas of Europe, all three predictions are upheld. Especiallysignificant is the occurrence, in close proximity, of a reconstructednorthern, circum-Arctic dispersal route (Martin et al., 2005 andreferences therein), of such clearly Asiamerican faunal elements asneoceratopsians and herpetotheriid marsupials. The presence of basallambeosaurines, suggesting faunal connections with Asian assem-blages during earlier Late Cretaceous times (?Santonian–Campanian;Pereda-Suberbiola et al., 2009; Prieto-Marquez and Wagner, 2009),within the Ibero-Armorican Realm, at the opposite end of Europe fromHaţeg, is intriguing and calls for the presence of a high-latitudedispersal route fromAsia towards continental Europe. It is conceivablethat this filter dispersal route was also responsible for the dispersal ofneoceratopsians into the more northern parts of Europe (as alsosuggested by Prieto-Marquez and Wagner, 2009). On the other hand,the presence of abelisauroids (possibly even relatively derivedabelisaurids; Carrano and Sampson, 2008), bothremydines (Rosasia;Gaffney et al., 2006) and relatively common madtsoiids, both cladeswith Gondwanan ties, on the Ibero-Armorican landmass, as well astheir occurrence near a suggested Iberian dispersal route connectingwith Africa (Gheerbrant and Rage, 2006) conforms again to predic-tions 2–3. The near-absence of all these groups in southeasternEurope (corresponding to Tethyan Europe, as opposed to cratonicEurope) suggests that the deep-sea trenches and basins transectingthis area represented barriers effective enough to restrict dispersal ofmany western European taxa.

Again, the Malay Archipelago can offer a parallel to this phenom-enon. The important biogeographic boundary dividing the archipelago(Wallace, 1860), called subsequently the Wallace Line, is presumably aby-product of the Lombok Strait, a deep-water trough separating theislands of Bali and Lombok. Although the nature, position, andeffectiveness of this dividing line is still under discussion (cf., Mayr,1944; Simpson, 1977; Brown and Guttman, 2002), biogeographicboundaries produced by deep-water barriers and affecting at leastsome continental vertebrate groupswithin anoceanic archipelago seemto be well established (Sweet and Pianka, 2003, 2007).

Similar to the present-day Malay Archipelago, biogeographicboundary lines dividing the European Archipelago (and especiallyits Tethyan segment) should be seen as the rule and not the exception.The peculiar composition of the Haţeg fauna, compared to those fromthe western part of Europe, supports the presence of such clear-cutboundaries within the “European palaeobioprovince”. This being thecase, it raises questions about the reality of a unique Europeanbioprovince, as hypothesized by several authors (e.g., Le Loeuff, 1997;Holtz et al., 2004; Pereda-Suberbiola, 2009). However, Late Creta-ceous Europe cannot be described as a Euro-African Province (Le

434 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Loeuff, 1991) either. Instead, the “European” bioprovince should beseen more as a mosaic within which the individuality of eachparticular faunal assemblage was shaped by different palaeobiogeo-graphic events and influences.

6. Conclusions

Faunal analyses of the European Cretaceous identify the Haţeg faunaas part of a larger European bioprovince. However, it remains somewhatatypical because of extreme species-level endemism compared tocontemporary terrestrial faunas from elsewhere in Europe. Phylogeneticanalyses offiveHaţeg species, calibrated by biostratigraphic occurrences,indicate long ghost lineages (30–80 M.y; mean=57), in keeping withthe dearth of preservation possibilities of insular faunas in the fossilrecord. Kogaionids,Kallokibotion,Allodaposuchus, and Zalmoxes (togetherwith their European sister taxa) are thought to have arisen fromvicariantevents between western Europe and North America. In contrast, thegeographic distribution of Telmatosaurus is an example of Europeanendemism following an Asiamerican ancestry and immigration.

While Haţeg likely acted as a refugium for Kallokibotion andTelmatosaurus, other faunal members (together with their immediatesister taxa) are not restricted to Transylvania, but known otherwisefrom localities across southern Europe. In addition, Transylvania mayhave acted as an evolutionary cradle for kogaionids.

Transylvania and the other southern faunas of Europe seem torepresent a distinct division of the Late Cretaceous Europeanpalaeobioprovince. A boundary between this Tethyan part of Europeand the more northern and western parts of the Europeanpalaeobioprovince would have functioned much like the WallaceLine in the Malay Archipelago, in which a seemingly homogeneousgeographic region is divided into two (or more) biogeographicprovinces that reflect different histories.

Acknowledgments

We are grateful to the many friends and colleagues (Attila Ősi,Márton Rabi — Budapest, Hungary; Xabier Pereda-Suberbiola —

Bilbao, Spain; Jean Le Loeuff — Espéraza, France; Jeremy E. Martin —

Corti, France; Francois Therrien — Drumheller, Alberta, Canada; MikeHabib— Baltimore, Maryland, USA; Cora Jianu— Deva, Romania) whowere either willing or inadvertent commentators on this study.Thanks are given to the CNCSIS (National University ResearchCouncil) grants to DG (1163A/2004) and ZCs (95At/2003, 1930/2009), as well as the Synthesys grant GB-TAF 3417-2007 to ZCs,supporting field andmuseum research. Thanks to the Royal Society forfunding the visit of ZCs to Bristol in 2006, through a RS InternationalIncoming Short Visit grant, offering the opportunity to outline severaltopics discussed in this contribution. DBW thanks the National ScienceFoundation, the National Geographic Society, the Dinosaur Society,the Jurassic Foundation, and the Paleontological Society InternationalResearch Program (the latter two awarded to Coralia-Maria Jianu,who is also thanked for her contributions to this paper) for fundssupporting field and museum research since 1987.

References

Apesteguía, S., 2002. Greater Gondwana and the Kawas Sea coastal tetrapod fauna(Campanian–Maastrichtian). II Simposio del Cretácico de América del Sur, SãoPedro, Brasil. 135–141.

Averianov, A.O., Yarkov, A.A., 2004. Carnivorous dinosaurs (Saurischia, Theropoda)from the Maastrichtian of the Volga-Don Interfluve, Russia. Paleontological Journal38, 78–82.

Badiola, A., Canudo, J.I., Cuenca-Bescós, G., 2008. New multituberculate mammals fromthe Hauterivian–Barremian transition of Europe (Iberian Peninsula). Palaeontology51, 1455–1469.

Barrett, P.M., Wang, X.-L., 2007. Basal titanosauriform (Dinosauria, Sauropoda) teethfrom the Lower Cretaceous Yixian Formation of Liaoning Province, China.Palaeoworld 16, 265–271.

Barrett, P.M., Hasegawa, Y., Manabe, M., Isaji, S., Matsouka, H., 2002. Sauropoddinosaurs from the Lower Cretaceous of eastern Asia: taxonomic and biogeo-graphical implications. Palaeontology 45, 1197–1217.

Bellemain, E., Ricklefs, R.E., 2008. Are islands the end of the colonization road? Trends inEcology and Evolution 23, 461–465.

Benton, M.J., Csiki, Z., Grigorescu, D., Redelstorff, R., Sander, P.M., Stein, K., Weishampel,D.B., 2010. Dinosaurs and the island rule: The dwarfed dinosaurs from Haţeg Island.Palaeogeography, Palaeoclimatology, Palaeoecology 293, 438–454 (this issue).

Bonaparte, J.F., 1984. Late Cretaceous faunal interchange of terrestrial vertebrates betweenthe Americas. In: Reif, W.-E., Westphal, F. (Eds.), Short Papers, Third Symposium onMesozoic Terrestrial Ecosystems. Attempo Verlag, Tübingen, pp. 19–24.

Bonaparte, J.F., Kielan-Jaworowska, Z., 1987. Late Cretaceous dinosaur and mammalfaunas of Laurasia and Gondwana. In: Currie, P.J., Koster, E.H. (Eds.), FourthSymposium on Mesozoic Terrestrial Ecosystems: Occasional Papers of the TyrellMuseum of Paleontology, 3. Drumheller, Alberta, pp. 24–29.

Brochu, C.A., 1999. Phylogeny, systematics, and historical biogeography of Alligator-oidea. Society of Vertebrate Paleontology Memoir 6, 9–100.

Brooks, D.R., McLennan, D.A., 1991. Phylogeny, Ecology, and Behavior. University ofChicago Press, Chicago. 434 pp.

Brown, R.M., Guttman, S.I., 2002. Phylogenetic systematics of the Rana signata complexof Philippine and Bornean stream frogs: reconsideration of Huxley's modification ofWallace's Line at the Oriental–Australian faunal zone interface. Biological Journal ofthe Linnean Society 76, 393–461.

Buffetaut, E., 1989. Archosaurian reptiles with Gondwanan affinities in the UpperCretaceous of Europe. Terra Nova 1, 69–74.

Buffetaut, E., Le Loeuff, J., 1991. Late Cretaceous dinosaur faunas of Europe: somecorrelation problems. Cretaceous Research 12, 159–176.

Buffetaut, E., Rage, J.-C., 1993. Fossil amphibians and reptiles and the Africa–SouthAmerica connection. In: George, W., Lavocat, R. (Eds.), The Africa–South Americaconnection. Clarendon Press, pp. 87–99.

Buffetaut, E., Le Loeuff, J., Tong, H., Duffaud, S., Cavin, L., Garcia, G., Ward, D.,l'Association culturelle, archeologique et paleontologique de Cruzy, 1999. Unnouveau gisement de vertébrés du Crétacé supérieur a Cruzy (Hérault, Sud de laFrance). Comptes Rendus de l'Académie des Sciences Paris 328, 203–208.

Buffetaut, E., Grigorescu, D., Csiki, Z., 2002. A new giant pterosaur with a robust skullfrom the latest Cretaceous of Romania. Naturwissenschaften 89, 180–184.

Buscalioni, A.D., Ortega, F., Vasse, D., 1999. The Upper Cretaceous crocodilianassemblage from Lano (northcentral Spain): implications in the knowledge ofthe fini-Cretaceous European faunas. Estudio Museo Ciencias Naturales de Alava14, 213–233.

Buscalioni, A.D., Ortega, F., Weishampel, D.B., Jianu, C.M., 2001. A revision of thecrocodyliform Allodaposuchus precedens from the Upper Cretaceous of the HaţegBasin, Romania. Its relevance in the phylogeny of Eusuchia. Journal of VertebratePaleontology 21, 74–86.

Butler, R.J., Upchurch, P., Norman, D.B., Parish, J., 2006. A biogeographical analysis of theornitischian dinosaurs. In: Barrett, P.M., Evans, S.E. (Eds.), Abstracts and Proceed-ings. Ninth International Symposium onMesozoic Terrestrial Ecosystems and Biota.Manchester, UK, pp. 13–16.

Butler, R.J., Upchurch, P., Norman, D.B., 2008. The phylogeny of the ornitischiandinosaurs. Journal of Systematic Palaeontology 6, 1–40.

Canudo, J.I., Ruiz-Omeñaca, J.I., Barco, J.L., Royo Torres, R., 2002. ¿Saurópods asiáticos enel Barremiense inferior (Cretácio Inferior) de Espaňa? Ameghiniana 39, 443–452.

Canudo, J.I., Barco, J.L., Pereda-Suberbiola, X., Ruiz-Omeñaca, J.I., Salgado, L., TorcidaFernández-Baldor, F., Gasulla, J.M., 2009. What Iberian dinosaurs reveal about thebridge said to exist between Gondwana and Laurasia in the Early Cretaceous.Bulletin. Société Géologique de France 180, 5–11.

Carpenter, K., 2001. Phylogenetic analysis of the Ankylosauria. In: Carpenter, K. (Ed.),The Armored Dinosaurs. Indiana University Press, Bloomington, pp. 455–483.

Carr, T.D., Williamson, T.E., Schwimmer, D.S., 2005. A new genus and species oftyrannosauroid from the Late Cretaceous (Middle Campanian) DemopolisFormation of Alabama. Journal of Vertebrate Paleontology 25, 119–143.

Carrano, M.T., Sampson, S.D., 2008. The phylogeny of Ceratosauria (Dinosauria:Theropoda). Journal of Systematic Palaeontology 6, 183–236.

Cifelli, R.L., 1997. First notice on Mesozoic mammals from Oklahoma. OklahomaGeology Notes, Oklahoma Geological Survey 57, 4–17.

Cifelli, R.L., 2000. Cretaceous mammals of Asia and North America. PaleontologicalSociety of Korea Special Publication 4, 49–84.

Clark, J.M., Norell, M.A., 1992. The Early Cretaceous crocodylomorph Hylaeochampsavectiana from the Wealden of the Isle of Wight. American Museum Novitates 3032,1–19.

Codrea, V.A., Dica, E.P., 2005. Upper Cretaceous–lowermost Miocene lithostratigraphicunits exposed in Alba Iulia – Sebeş – Vinţu de Jos area (SW Transylvanian basin).Studia Universitatis Babeş-Bolyai, Geologia 50, 19–26.

Codrea, V., Smith, T., Dica, P., Folie, A., Garcia, G., Godefroit, P., Van Itterbeeck, J., 2002.Dinosaur egg nests, mammals and other vertebrates from a newMaastrichtian siteof the Haţeg Basin (Romania). Comptes Rendus Palevol 1, 173–180.

Codrea, V., Godefroit, P., Smith, T., Jipa-Murzea, C., 2009a. Maastrichtian landvertebrates in Rusca Montană Basin (Romania). In: Godefroit, P., Lambert, O.(Eds.), Tribute to Charles Darwin and Bernissart Iguanodons: New Perspectives onVertebrate Evolution and Early Cretaceous Ecosystems, p. 29. Brussels.

Codrea, V., Vremir, M., Jipa, C., Godefroit, P., Csiki, Z., Smith, T., Fărcaş, C., 2010. Morethan just Nopcsa's Transylvanian dinosaurs: a look outside the Haţeg Basin.Palaeogeography, Palaeoclimatology, Palaeoecology 293, 391–405 (this issue).

Csiki, Z., 2005. Sistematica, tafonomia şi paleoecologia microvertebratelor şi dinosaur-ilor saurischieni din Maastrichtianul Bazinului Haţeg. Unpubl. Ph.D. Thesis,University of Bucharest, 527 pp. [In Romanian]

435D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Csiki, Z., Grigorescu, D., 2000. Teeth of multituberculate mammals from the LateCretaceous of Romania. Acta Palaeontologica Polonica 45, 85–90.

Csiki, Z., Grigorescu, D., 2001. Fossil mammals from the Maastrichtian of the HaţegBasin, Romania. Abstracts Volume, 6th European Workshop on VertebratePaleontology, Florence, Italy, p. 26.

Csiki, Z., Grigorescu, D., 2002. Paleobiogeographical implications of the fossil mammalsfrom the Maastrichtian of the Haţeg Basin. Acta Palaeontologica Romaniae 3, 87–96.

Csiki, Z., Grigorescu, D., 2005. A new theropod from Tustea: are there oviraptorosaurs inthe Upper Cretaceous of Europe? Abstracts volume, 3rd Annual Meeting of theEuropean Association of Vertebrate Palaeontologists. Darmstadt, Germany. 78.

Csiki, Z., Grigorescu, D., 2006. Maastrichtian multituberculates of the Haţeg Basin,Romania: implications for multituberculate phylogeny, evolution and paleobio-geography. In: Barrett, P.M., Evans, S.E. (Eds.), Abstracts and Proceedings. NinthInternational Symposium on Mesozoic Terrestrial Ecosystems and Biota, Manche-ster, UK, pp. 29–32.

Csiki, Z., Grigorescu, D., 2007. The “Dinosaur Island” – new interpretation of the Haţegvertebrate fauna after 110 years. Sargetia 20, 5–26.

Csiki, Z., Grigorescu, D., Rücklin, M., 2005. A new multituberculate specimen from theMaastrichtian of Pui, Romania and reassessment of affinities of Barbatodon. ActaPalaeontologica Romaniae 5, 73–86.

Csiki, Z., Grigorescu, D., Weishampel, D.B., 2007. A new titanosaur sauropod(Dinosauria: Saurischia) from the Upper Cretaceous of the Haţeg Basin (Romania).Abstracts volume, 5th Annual Meeting of the European Association of VertebratePalaeontologists (EAVP), Carcassone, France, p. 17.

Csiki, Z., Codrea, V., Jipa-Murzea, C., Godefroit, P., in press. A partial titanosaur(Sauropoda, Dinosauria) skeleton from the Maastrichtian of Nălaţ-Vad, HaţegBasin, Romania. Neues Jahrbuch für Geologie und Paläontologie.

Csontos, L., Vörös, A., 2004. Mesozoic plate tectonic reconstruction of the Carpathianregion. Palaeogeography, Palaeoclimatology, Palaeoecology 210, 1–56.

Cuenca-Bescós, G., Canudo, J.I., 2003. A new gobiconodontid mammal from the EarlyCretaceous of Spain and its palaeogeographic implications. Acta PalaeontologicaPolonica 48, 575–582.

Curry Rogers, K., 2005. The evolutionary history of the Titanosauria. In: Curry Rogers, K.,Wilson, J.A. (Eds.), The Sauropods: Evolution and Paleobiology. University ofCalifornia Press, Berkeley, pp. 50–103.

Dalla Vecchia, F., 2006. Telmatosaurus and the other hadrosaurids of the CretaceousEuropean archipelago. An overview, Natura Nascosta 32, 1–55.

Danilov, I.G., Parham, J.F., 2008. A reassessment of some poorly known turtles from theMiddle Jurassic of China, with comments on the antiquity of extant turtles. Journalof Vertebrate Paleontology 28, 306–318.

Delfino, M., Codrea, V., Folie, A., Dica, P., Godefroit, P., Smith, T., 2008. A complete skullof Allodaposuchus precedens Nopcsa, 1928 (Eusuchia) and a reassessment of themorphology of the taxon based on the Romanian remains. Journal of VertebratePaleontology 28, 111–122.

Denton, R.K., O'Neil, R.C., Grandstaff, B., Parris, D.C., 1996. The Ellisdale dinosaur site: anoverview of its geology and paleontology. In: Gallagher, W.B., Parris, D.C. (Eds.),Cenozoic and Mesozoic vertebrate paleontology of the New Jersey coastal plain;SVP fieldtrip, pp. 19–34.

Dercourt, J., Gaetani, M., Vrielynck, B., Barrier, E., Biju-Duval, B., Brunet, M., Cadet, J. P.,Crasquin, S., Săndulescu, M. (Eds.), 2000. Atlas Peri-Tethys, PalaeogeographicalMaps. CCGM/CGMW, Paris, 269 pp., 24 maps.

Estes, R., Sanchíz, B., 1982. Early Cretaceous lower vertebrates from Galve (Teruel),Spain. Journal of Vertebrate Paleontology 2, 21–29.

Farris, J.S., 1970. Methods for computing Wagner trees. Systematic Zoology 19, 83–92.Folie, A., Codrea, V., 2005. New lissamphibians and squamates from theMaastrichtian of

Haţeg Basin, Romania. Acta Palaeontologica Polonica 50, 57–71.Forster, C.A., 1997. Hadrosauridae. In: Currie, P.J., Padian, K. (Eds.), Encyclopedia of

Dinosaurs. Academic Press, San Diego, pp. 293–299.Forster, C.A., 1999. Gondwanan dinosaur evolution and biogeographic analysis. Journal

of African Earth Sciences 28, 169–185.Gaffney, E.S., Meylan, P.A., 1992. The Transylvanian turtle, Kallokibotion, a primitive

cryptodire of Cretaceous age. American Museum Novitates 3040, 1–37.Gaffney, E.S., Tong, H., Meylan, P.A., 2006. Evolution of the side-necked turtles: the

families Bothremydidae, Euraxemydidae, and Araripemydidae. Bulletin of theAmerican Museum of Natural History 300, 1–698.

Garcia, G., Pereda-Suberbiola, X., 2003. A new species of Struthiosaurus (Dinosauria:Ankylosauria) from the Upper Cretaceous of Villeveyrac (southern France). Journalof Vertebrate Paleontology 23, 156–165.

Gayet, M., Rage, J.-C., Sempere, Th., Gagnier, P.-Y., 1992. Modalités des échanges desvertébres continentaux entre l'Amérique du Nord et l'Amérique du Sud au Crétacésupérieur et au Paléocène. Bulletin. Société Géologique de France 163, 781–791.

Gheerbrant, E., Rage, J.C., 2006. Paleobiogeography of Africa: how distinct fromGondwana and Laurasia? Palaeogeography, Palaeclimatology, Palaeoecology 241,224–246.

Gheerbrant, E., Codrea, V., Hosu, A., Sen, S., Guernet, C., de Lapparent de Broin, F.,Riveline, J., 1999. Découverte en Transylvanie (Roumanie) de gisements a vertébresdans les Calcaires de Rona (Thanétien ou Sparnacien): le plus anciens mammiferescénozoiques d'Europe Orientale. Eclogae Geologica Helvetiae 92, 517–535.

Godefroit, P., Lambert, O., 2007. A re-appraisal of Craspedodon lonzeensis Dollo, 1883,from the Upper Cretaceous of Belgium: the first record of a neoceratopsiandinosaur in Europe? Bulletin de l'Institut Royal des Sciences Naturelles de Belgique77, 83–93.

Godefroit, P., Dong, Z.-M., Bultynck, P., Li, H., Feng, L., 1998. New Bactrosaurus(Dinosauria: Hadrosauroidea) material from Iren Dabasu (Inner Mongolia, PRChina). Bulletin de l'Institut royal des Sciences naturelles de Belgique, Sciences de laTerre 68, 3–70 Supplement.

Godefroit, P., Bolotsky, Y.L., Alifanov, V., 2003. A remarkable hollow-crested hadrosaurfrom Russia: an Asian origin for lambeosaurines. Comptes Rendus Palevol 2,143–151.

Godefroit, P., Bolotsky, Y.L., Van Itterbeeck, J., 2004. The lambeosaurine dinosaurAmurosaurus riabinini, from the Maastrichtian of Far Eastern Russia. ActaPalaeontologica Polonica 49, 585–618.

Golonka, J., 2004. Plate tectonic evolution of the southern margin of Eurasia in theMesozoic and Cenozoic. Tectonophysics 381, 235–273.

Grande, L., 1990. Vicariance biogeography. In: Briggs, D.E.G., Crowther, P.R. (Eds.),Palaeobiology. Blackwell Scientific Publications, A synthesis, pp. 444–451.

Grigorescu, D., 2005. Rediscovery of a “forgotten land”. The last three decades ofresearch on the dinosaur-bearing deposits from the Haţeg Basin. Acta Palaeonto-logica Romaniae 5, 191–204.

Grigorescu, D., 2010. The Latest Cretaceous fauna with dinosaurs and mammals fromthe Haţeg Basin — A historical overview. Palaeogeography, Palaeoclimatology,Palaeoecology 293, 271–282 (this issue).

Grigorescu, D., Hartenberger, J.-L., Rădulescu, C., Samson, P., Sudre, J., 1985. Découvertede mammifères et de dinosaures dans le Crétacé supérieur de Pui (Roumanie).Comptes rendus de l'Académie des Sciences de Paris 301, 1365–1368.

Grigorescu, D., Venczel, M., Csiki, Z., Limberea, R., 1999. New latest Cretaceousmicrovertebrate fossil assemblages from the Haţeg Basin (Romania). Geologie enMijnbouw 98, 310–314.

Guthrie, D., 2004. Radiocarbon evidence of mid-Holocene mammoth stranded on anAlaskan Bering Sea island. Nature 429, 746–749.

Hahn, G., Hahn, R., 2002. Neue Multituberculaten-Zähne aus dem Barremium (Unter-Kreide) von Galve (Spanien). Paläontologische Zeitschrift 76, 257–259.

Hay, W.W., DeConto, R.M., Wold, C.N., Wilson, K.M., Voigt, S., Schulz, M., Wold-Rossby,A., Dullo, W.-C., Ronov, A.B., Balukhovsky, A.N., Söding, E., 1999. Alternative globalCretaceous paleogeography. In: Barrera, E., Johnson, C.C. (Eds.), Evolution of theCretaceous Ocean–Climate System: Geological Society of America Special Paper,332, pp. 1–47.

Head, J.J., 1998. A new species of basal hadrosaurid (Dinosauria, Ornithischia) from theCenomanian of Texas. Journal of Vertebrate Paleontology 18, 718–738.

Hirayama, R., Brinkman, D.B., Danilov, I.G., 2000. Distribution and biogeography of non-marine Cretaceous turtles. Russian Journal of Herpetology 7, 181–198.

Holtz, T.R., Chapman, R.E., Lamanna, M.C., 2004. Mesozoic biogeography of Dinosauria,In: Weishampel, D.B., Dodson, P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition.University of California Press, Berkeley, Los Angeles and London, pp. 627–642.

Horner, J.R., Weishampel, D.B., Forster, C.A., 2004. Hadrosauridae, In: Weishampel, D.B.,Dodson, P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition. University ofCalifornia Press, Berkeley, Los Angeles and London, pp. 438–463.

Hu, Y.M., Wang, Y.Q., 2002. Sinobaatar gen nov.: first multituberculate from the JeholBiota of Liaoning, Northeast China. Chinese Science Bulletin 47, 933–938.

Jablonski, D., 1986. Causes and consequences of mass extinctions. In: Elliot, D.K. (Ed.),Dynamics of Extinction. John Wiley, New York, pp. 183–229.

Jianu, C.-M., Boekschoten, G.J., 1999. The Haţeg — island or outpost? Deinsea 7,195–198.

Jianu, C.-M., Weishampel, D.B., Ştiucă, E., 1997. Old and new pterosaur material fromthe Haţeg Basin (Late Cretaceous) of western Romania and comments aboutpterosaur diversity in the Late Cretaceous of Europe. Second European Workshopon vertebrate palaeontology (Espéraza — Quillan), Abstracts. Musée des Dino-saures, Espéraza, France.

Jones, M.E.H., Tennyson, A.J.D., Worthy, J.P., Evans, S.E., Worthy, T.H., 2009. Asphenodontine (Rhynchocephalia) from the Miocene of New Zealand andpalaeobiogeography of the tuatara (Sphenodon). Proceedings of the Royal SocietyB 276, 1385–1390.

Joyce, W.G., 2007. Phylogenetic relationships of Mesozoic turtles. Bulletin of thePeabody Museum of Natural History 48, 3–102.

Kessler, E., Grigorescu, D., Csiki, Z., 2005. Elopteryx revisited— a new bird-like specimenfrom the Maastrichtian of the Haţeg Basin (Romania). Acta PalaeontologicaRomaniae 5, 249–258.

Kielan-Jaworowska, Z., Hurum, J.H., 1997. Djadochtatheria — a new suborder ofmultituberculate mammals. Acta Palaeontologica Polonica 42, 201–242.

Kielan-Jaworowska, Z., Hurum, J.H., 2001. Phylogeny and systematics of multi-tuberculate mammals. Palaeontology 44 (3), 389–429.

Kielan-Jaworowska, Z., Dashzeveg, D., Trofimov, B.A., 1987. Early Cretaceous multi-tuberculates from Mongolia and a comparison with Late Jurassic forms. ActaPalaeontologica Polonica 32, 3–47.

Kielan-Jaworowska, Z., Cifelli, R.L., Luo, Z.X., 2004. Mammals from the age of dinosaurs:origins, evolution and structure. Columbia University Press, New York. 630 pp.

Kirkland, J.I., 1998. A new hadrosaurid from the Upper Cretaceous Cedar MountainFormation (Albian–Cenomanian: Cretaceous) of eastern Utah — the oldest knownhadrosaurid (Lambeosaurinae?). In: Lucas, S.G., Kirkland, J.I., Estep, J.W. (Eds.),Lower and Middle Cretaceous Terrestrial Ecosystems: New Mexico Museum ofNatural History and Science Bulletin, 14, pp. 283–295.

Kirkland, J.I., Britt, B., Burge, D.L., Carpenter, K., Cifelli, R.L., DeCourten, F., Eaton, J., Hasiotis,S., Lawton, T., 1997. Lower to middle Cretaceous dinosaur faunas of the centralColorado Plateau: a key to understanding 35 M.y. of tectonics, sedimentology,evolution, and biogeography. Brigham Young University Geology Studies 42, 69–103.

Krause, D.W., O'Connor, P.M., Curry Rogers, K., Sampson, S.D., Buckley, G.A., Rogers, R.R.,2006. Late Cretaceous terrestrial vertebrates from Madagascar: implications forLatin American biogeography. Annals of the Missouri Botanical Garden 93,178–208.

Kusuhashi, N., 2008. Early Cretaceous multituberculate mammals from the KuwajimaFormation (Tetori Group), central Japan. Acta Palaeontologica Polonica 53,379–390.

436 D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Lapparent, A.F., 1947. Les dinosauriens du Crétacé supérieur du midi de la France.Mémoire de la Société géologique de France 56, 1–54.

Lapparent de Broin, F., Murelaga, X., 1999. Turtles from the Upper Cretaceous of Lano(Iberian Peninsula). Estudios Museo Ciencias Naturales de Alava 14, 135–211.

Laurent, Y., Le Loeuff, J., Buffetaut, E., 1997. Les Hadrosauridae (Dinosauria,Ornithopoda) du Maastrichtien supérieur des Corbières orientales (Aude, France).Revue de Paléobiologie 16, 411–423.

Le Loeuff, J., 1991. The Campano–Maastrichtian vertebrate faunas from SouthernEurope and their relationships with other faunas in the world: Palaeobiogeogra-phical implications. Cretaceous Research 12, 93–114.

Le Loeuff, J., 1993. European titanosaurs. Revue de Paléobiologie 7, 105–117.Le Loeuff, J., 1997. Biogeography. In: Currie, P.J., Padian, K. (Eds.), Encyclopedia of

Dinosaurs. Academic Press, San Diego, pp. 51–56.Le Loeuff, J., Buffetaut, E., 1995. The evolution of the Late Cretaceous non-marine vertebrate

fauna in Europe. In: Sun, A., Wang, Y. (Eds.), Sixth Symposium onMesozoic TerrestrialEcosystems and Biota. Short Papers. China Ocean Press, Beijing, pp. 181–184.

Le Loeuff, J., Buffetaut, E., Martin, M., Martin, V., Tong, H., 1993. Découverted'Hadrosauridae (Dinosauria, Ornithischia) dans le Maastrichtien des Corbières(Aude, France). Comptes Rendus de l'Académie des Sciences Paris 316, 1023–1029.

Lindgren, J., Currie, P.J., Siverson, M., Rees, J., Cederström, P., Lindgren, F., 2007. The firstneoceratopsian dinosaur remains from Europe. Palaeontology 50, 929–937.

Longrich, N.R., Currie, P.J., 2009. Albertonykus borealis, a new alvarezsaur (Dinosauria:Theropoda) from the Early Maastrichtian of Alberta, Canada: implications forthe systematics and ecology of theAlvarezsauridae. Cretaceous Research 30, 239–252.

Lull, R.S., Wright, N.E., 1942. Hadrosaurian dinosaurs of North America. GeologicalSociety of America Special Paper 40, 1–242.

Lydekker, R., 1888. Note on a new Wealden iguanodont and other dinosaurs. QuarterlyJournal Geological Society of London 44, 46–61.

Makádi, L., 2006. Bicuspidon aff. hatzegiensis (Squamata: Scincomorpha: Teiidae) fromthe Upper Cretaceous Csehbánya Formation (Hungary, Bakony Mts). ActaGeologica Hungarica 49, 373–385.

Makovicky, Martin J.E., 2007. New material of the Late Cretaceous globidontanAcynodon iberoccitanus (Crocodylia) from the Late Cretaceous of southern France.Journal of Vertebrate Paleontology 27, 362–372.

Martin, J.E., 2010. Allodaposuchus Nopcsa, 1928 (Crocodylia, Eusuchia) from the LateCretaceous of southern France and its relationships to Alligatoroidea. Journal ofVertebrate Paleontology 30, 756–767.

Martin, J.E., Buffetaut, E., 2005. An overview of the Late Cretaceous crocodilianassemblage from Cruzy, southern France. Kaupia 14, 33–40.

Martin, J.E., Buffetaut, E., 2008. Crocodilus affuvelensis Matheron, 1869 from the LateCretaceous of southern France: a reassessment. Zoological Journal of the LinnaeanSociety 152, 567–580.

Martin, J.E., Delfino, M., 2010. Recent advances on the comprehension of thebiogeography of Cretaeous European eusuchians. Palaeogeography, Palaeoclima-tology, Palaeoecology 293, 406–418 (this issue).

Martin, J.E., Case, J.A., Jagt, J.W.M., Schulp, A.S., Mulder, E.W.A., 2005. A new Europeanmarsupial indicates a Late Cretaceous high-latitude transatlantic dispersal route.Journal of Mammalian Evolution 12, 495–510.

Martin, J.E., Csiki, Z., Grigorescu, D., Buffetaut, E., 2006. Late Cretaceous crocodiliandiversity in Haţeg Basin, Romania. Hantkeniana 5, 31–37.

Matheron, M.P., 1869. Notice sur les Reptiles fossiles des dépots fluvio-lacustresCrétacés du bassin a lignite de Fuveau. Mémoire de l'Academie Imperiale desSciences, Belles-Lettres et Arts de Marseille. 1–39.

Matzke, A.T., Maisch, M.W., Sun, G., Pfretzschner, H.U., Stöhr, H., 2005. A new MiddleJurassic xinjiangchelyid turtle (Testudines; Eucryptodira) from China (Xinjiang,Junggar Basin). Journal of Vertebrate Paleontology 25, 63–70.

Mayr, E., 1944.Wallace's Line in light of recent zoogeographic studies. Quarterly Reviewof Biology 19, 1–14.

Milner, A.R., Norman, D.B., 1984. The biogeography of advanced ornithopod dinosaurs(Archosauria: Ornithischia) — a cladistic–vicariance model. In: Reif, W.E.,Westphal, F. (Eds.), Third Symposium on Mesozoic Terrestrial Ecosystems. ShortPapers. Attempto Verlag, Tübingen, pp. 145–151.

Mitchell, N.J., Nelson, N.J., Cree, A., Pledger, S., Keall, S., Daugherty, C.H., 2006. Supportfor a rare pattern of temperature-dependent sex determination in archaic reptiles:evidence from two species of tuatara (Sphenodon). Frontiers in Zoology 3, 9.doi:10.1186/1742-9994-3-9.

Molnar, R.E., 1980. Australian late Mesozoic terrestrial tetrapods: some implications.Mémoires de la Societé Géologique de France 139, 131–143.

Naish, D., Dyke, G.J., 2004. Heptasteornis was no ornithomimid, troodontid, dromaeo-saurid or owl: the first alvarezsaurid (Dinosauria: Theropoda) from Europe. NeuesJahrbuch für Geologie und Paläontologie Monatshefte 2004, 385–401.

Nekola, J.C., 1999. Paleorefugia and neorefugia: the influence of colonization history oncommunity pattern and process. Ecology 80, 2459–2473.

Nelson, G., Platnick, N., 1981. Systematics and Biogeography. Cladistics and Vicariance.Columbia University Press, New York. 567 pp.

Newton, C.R., 1990. Palaeobiogeography. In: Briggs, D.E.G., Crowther, P.R. (Eds.),Palaeobiology. A Synthesis, Blackwell Scientific Publications, pp. 452–461.

Nopcsa, F., 1900. Dinosaurierreste aus Siebenbürgen I. Schädel von Limnosaurustranssylvanicus nov. gen. et nov. spec. Denkschriften der königlichen Akademie derWissenschaften: Mathematisch-Naturwissenschaftlichen Klasse, 68, pp. 555–591.

Nopcsa, F., 1902. Dinosaurierreste aus Siebenbürgen. II. Schädelreste von Mochlodon.Mit einem Anhange: zur Phylogenie der Ornithopodiden. Denkschriften derköniglichen Akademie der Wissenschaften: Mathematisch-Naturwissenschaftli-chen Klasse, 72, pp. 149–175.

Nopcsa, 1903. Telmatosaurus, a new name for the dinosaur Limnosaurus. GeologicalMagazine (Series 5) 2, 241–295.

Nopcsa, F., 1915. Die Dinosaurier der siebenbürgischen landesteile Ungarns. Mitteilun-gen aus dem Jahrbuche der KGL. Ungarischen Geologischen Reichsanstalt, pp. 1–24.

Nopcsa, F., 1923a. On the geological importance of the primitive reptilian fauna of theuppermost Cretaceous of Hungary; with a description of a new tortoise(Kallokibotium). Quarterly Journal Geological Society of London 79, 100–116.

Nopcsa, F., 1923b. Kallokibotium, a primitive amphichelydean tortoise from theuppermost Cretaceous of Hungary. Palaeontologica Hungarica 1, 1–34.

Nopcsa, F., 1928a. Paleontological notes on Reptilia. 7. Classification of the Crocodilia.Geologica Hungarica. Series Palaeontologica 1, 75–84.

Nopcsa, F., 1928b. Dinosaurierreste aus Siebenbürgen. IV. Die Wirbelsäule vonRhabdodon und Orthomerus. Palaeontologica Hungarica 1, 273–304.

Nopcsa, F., 1929. Dinosaurierreste aus Siebenbürgen. V. Geologica Hungarica. SeriesPaleontologica 4, 1–76.

Nopcsa, F., 1934. The influence of geological and climatological factors on thedistribution of non-marine fossil reptiles and Stegocephalia. Quarterly JournalGeological Society of London 90, 76–140.

Norell, M.A., 1992. Taxic origin and temporal diversity: the effects of phylogeny. In:Novacek, M.J., Wheeler, Q. (Eds.), Extinction and Phylogeny. Columbia UniversityPress, New York, pp. 89–118.

Norman, D.B., 2002. OnAsian ornithopods (Dinosauria: Ornithischia). 4. ProbactrosaurusRozhdestvensky, 1966. Zoological Journal of the Linnean Society 136, 113–144.

Norman, D.B., 2004. Basal Iguanodontia, In: Weishampel, D.B., Dodson, P., Osmólska, H.(Eds.), The Dinosauria, 2nd edition. University of California Press, Berkeley, LosAngeles and London, pp. 413–437.

Ősi, A., 2005. Hungarosaurus tormai, a new ankylosaur (Dinosauria) from the UpperCretaceous of Hungary. Journal of Vertebrate Paleontology 25, 370–383.

Ősi, A., Makádi, L., 2009. New remains of Hungarosaurus tormai (Ankylosauria,Dinosauria) from the Upper Cretaceous of Hungary: skeletal reconstruction andbody mass estimation. Paläontologische Zeitschrift 83, 227–245.

Ősi, A., Clark, J.M., Weishampel, D.B., 2007. First report on a new basal eusuchiancrocodyliform with multicusped teeth from the Upper Cretaceous (Santonian) ofHungary.Neues Jahrbuch fürGeologieundPaläontologieAbhandlungen243, 169–177.

Osmólska, H., Currie, P.J., Barsbold, R., 2004. Oviraptorosauria, In: Weishampel, D.B.,Dodson, P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition. University ofCalifornia Press, Berkeley, Los Angeles and London, pp. 165–183.

Ostrom, J.H., 1961. Cranial morphology of the hadrosaurian dinosaurs of North America.Bulletin of the American Museum of Natural History 122, 33–186.

Padian, K., 2004. Basal Avialae, In: Weishampel, D.B., Dodson, P., Osmólska, H. (Eds.),The Dinosauria, 2nd edition. University of California Press, Berkeley, Los Angelesand London, pp. 210–231.

Peláez-Campomanes, P., López-Martinez, N., Álvarez-Sierra, M.A., Daams, R., 2000. Theearliest mammal of the European Paleocene: the multituberculate Hainina. Journalof Paleontology 74, 701–711.

Peng, J.-H., Brinkman, D.B., 1993. New material of Xinjiangchelys (Reptilia: Testudines)from the Late Jurassic Qigu Formation (Shishugou Group) of the Pingfengshanlocality, Junggar Basin, Xinjiang. Canadian Journal of Earth Sciences 30,2013–2026.

Pereda-Suberbiola, X., 2009. Biogeographical affinities of Late Cretaceous continentaltetrapods of Europe: a review. Bulletin. Société Géologique de France 180, 57–71.

Pereda-Suberbiola, X., Galton, P.M., 2001. Reappraisal of the nodosaurid ankylosaurStruthiosaurus austriacus Bunzel from the Upper Cretaceous Gosau Beds of Austria. In:Carpenter, K. (Ed.), The Armored Dinosaurs. Indiana University Press, Bloomington,pp. 173–210.

Pereda-Suberbiola, X., Galton, P.M., 2009. Dwarf dinosaurs in the latest Cretaceous ofEurope? In: Colectivo Arqueológico-Paleontologógico Salense (Ed.), Actas, IVJornadas Internacionales sobre Paleontología de Dinosaurios y su Entorno. Salasde los Infantes (Burgos, Spain), pp. 263–272.

Pereda-Suberbiola, X., Canudo, J.I., Cruzado-Caballero, P., Barco, J.L., López-Martínez, N.,Oms, O., Ruiz-Omeňaca, J.I., 2009. The last hadrosaurid dinosaurs of Europe: a newlambeosaurine from the Uppermost Cretaceous of Aren (Huesca, Spain). ComptesRendus Palevol 8, 559–572.

Prasad, G.V.R., Rage, J.C., 1991. A discoglossid frog in the Latest Cretaceous(Maastrichtian) of India. Further evidence for a terrestrial route between Indiaand Laurasia in the Latest Cretaceous. Comptes rendus de l'Académie des Sciencesde Paris 313, 272–278.

Prieto-Marquez, A., Wagner, J.R., 2009. Pararhabdodon isonensis and Tsintaosaurusspinorhinus: a new clade of lambeosaurine hadrosaurids from Eurasia. CretaceousResearch 30, 1238–1246.

Rabi, M., 2005. Alligatoroidea indet. from the Upper Cretaceous of Hungary (Csehbányaformation). Kaupia 14, 93A.

Rădulescu, C., Samson, P.-M., 1986. Précisions sur les affinites des Multituberculés(Mammalia) du Crétacé supérieur de Roumanie. Comptes Rendus de l’Académiedes sciences Paris 304, 1825–1830.

Rădulescu, C., Samson, P.-M., 1996. The first multituberculate skull from the LateCretaceous (Maastrichtian) of Europe (Haţeg Basin, Romania). Anuarul InstitutuluiGeologic Român 69, 177–178.

Rage, J.-C., 2002. The continental Late Cretaceous of Europe: toward a betterunderstanding. Comptes Rendus Palevol 1, 257–258.

Rage, J.-C., 2003. Relationships of the Malagasy fauna during the Late Cretaceous:northern or southern routes? Acta Palaeontologica Poloniae 48, 661–662.

Rauhut, O.W.M., 2002. Dinosaur teeth from the Barremian of Uña, Province of Cuenca,Spain. Cretaceous Research 23, 255–263.

Rauhut, O.W.M., López-Arbarello, A., 2008. Archosaur evolution during the Jurassic: asouthern perspective. Revista de la Asociación Geológica Argentina 63, 557–585.

Reynoso, V.H., 2000. An unusual aquatic sphenodontian (Reptilia: Diapsida) from theTlayua Formation (Albian), central Mexico. Journal of Paleontology 74, 133–148.

437D.B. Weishampel et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 293 (2010) 419–437

Rougier, G.W., Novacek, M.J., Dashzeveg, D., 1997. A newmultituberculate from the LateCretaceous locality Ukhaa Tolgod. Mongolia. Considerations on multituberculaterelationships. American Museum Novitates 3193, 1–26.

Russell, D.A., 1993. The role of Central Asia in dinosaurian biogeography. CanadianJournal of Earth Sciences 30, 2002–2012.

Sachs, S., Hornung, J.J., 2006. Juvenile ornithopod (Dinosauria: Rhabdodontidae)remains from the Upper Cretaceous (Lower Campanian, Gosau Group) ofMuthmannsdorf (Lower Austria). Geobios 39, 415–425.

Salisbury, S.W., Frey, E., Martill, D.M., Buchy, M.-C., 2003. A new mesosuchiancrocodilian from the Lower Cretaceous Crato Formation of north-eastern Brazil.Palaontographica Abteilung 270, 3–47.

Salisbury, S.W., Molnar, R.E., Frey, E., Willis, P.M.A., 2006. The origin of moderncrocodyliforms: new evidence from the Cretaceous of Australia. Proceedings of theRoyal Society of London B 273, 2439–2448.

Sankey, J., Brinkman, D., Guenther, M., Currie, P., 2002. Small theropod and bird teethfrom the Late Cretaceous (Late Campanian) Judith River Group, Alberta. Journal ofVertebrate Paleontology 76, 751–763.

Schmid, S.M., Bernoulli, D., Fügenschuh, B.,Matenco, L., Schefer, S., Schuster, R., Tischler,M.,Ustaszewski, K., 2008. The Alpine–Carpathian–Dinaridic orogenic system: correlationand evolution of tectonic units. Swiss Journal of Geosciences 101, 139–183.

Schwimmer, D.R., 1997. Late Cretaceous dinosaurs in Eastern USA: a taphonomic andbiogeographic model of occurrences. In: Wolberg, D.L., Stump, E., Rosenberg, G.D.(Eds.), Dinofest International. Academy of Natural Sciences, pp. 203–211.

Scotese, C.R., 2004. A continental drift flipbook. Journal of Geology 112, 729–741.Sereno, P.C., 1999. Dinosaurian biogeography: vicariance, dispersal and regional

extinction. In: Tomida, Y., Rich, T.H., Vickers-Rich, P. (Eds.), Proceedings ofthe Second Gondwanan Dinosaur Symposium. National Science Museum, Tokyo,pp. 249–257.

Sereno, P.C., 2000. The fossil record, systematics and evolution of pachycephalosaursand ceratopsians from Asia. In: Benton, M.J., Shishkin, M.A., Unwin, D.M.,Kurochkin, E.N. (Eds.), The Age of Dinosaurs in Russia and Mongolia. CambridgeUniversity Press, pp. 480–516.

Sereno, P.C., Wilson, J.A., Conrad, J.L., 2004. New dinosaurs link southern landmasses inthe mid-Cretaceous. Proceedings of the Royal Society of London B 271, 1325–1330.

Simpson, G.G., 1977. Too many lines; the limits of the Oriental and Australianzoogeographic regions. Proceedings. American Philosophical Society 121, 107–119.

Smith, T., Codrea, V., 2003. Newmultituberculate mammals from the Late Cretaceous ofTransylvania (Romania). Abstracts Volume Fourth National Symposium ofPaleontology, Cluj-Napoca. pp. 51.

Smith, A.G., Smith, D.G., Funnell, B.M., 1994. Atlas of Mesozoic and Cenozoic Coastlines.Cambridge University Press, Cambridge. 204 pp.

Smith, T., Codrea, V., Săsăran, E., Van Itterbeck, J., Bultynck, P., Csiki, Z., Dica, P., Fărcaş, C.,Folie, A., Garcia, G., Godefroit, P., 2002. A new exceptional vertebrate site from theLate Cretaceous of the Haţeg Basin (Romania). Studia Universitatis Babeş-Bolyai,Geologia Special issue 1, 321–330.

Smith, N.D., Makovicky, P., Angolin, F.L., Ezcurra, M.D., Pais, D.F., Salisbury, S., 2008. AMegaraptor-like theropod (Dinosauria: Tetanurae) in Australia: support for faunalexchange across eastern and western Gondwana in the Mid-Cretaceous. Proceed-ings of the Royal Society of London B 275, 2085–2093.

Stampfli, G.M., Borel, G.D., 2002. A plate tectonic model for the Paleozoic and Mesozoicconstrained by dynamic plate boundaries and restored synthetic oceanicisochrones. Earth and Planetary Science Letters 196, 17–33.

Sues, H.-D., Averianov, A., 2009. A new basal hadrosauroid dinosaur from the LateCretaceous of Uzbekistan and the early radiation of duck-billed dinosaurs.Proceedings of the Royal Society of London B 276, 2549–2555.

Sukhanov, V.B., 2001. Mesozoic turtles from Mongolia. In: Benton, M.J., Shishkin, M.A.,Kurochkin, E.N., Unwin, D.M. (Eds.), The Age of Dinosaurs in Russia and Mongolia.Cambridge University Press, Cambridge, pp. 309–367.

Sullivan, R.M., 1999. Nodocephalosaurus kirtlandensis, gen et sp. nov., a newankylosaurid dinosaur (Ornitischia, Ankylosauria) from the Upper CretaceousKirtland Formation (Upper Campanian), San Juan Basin, New Mexico. Journal ofVertebrate Paleontology 19, 126–139.

Sweet, S.S., Pianka, E.R., 2003. The lizard kings. Natural History 112, 40–45.Sweet, S.S., Pianka, E.R., 2007. Monitors, mammals and Wallace's Line. Mertensiella 16,

79–99.Therrien, F., 2005. Palaeoenvironments of the latest Cretaceous (Maastrichtian)

dinosaurs of Romania: insights from fluvial deposits and paleosols of theTransylvanian and Haţeg basins. Palaeogeography, Palaeoclimatology, Palaeoecol-ogy 218, 15–56.

Turner, A.H., 2004. Crocodyliform biogeography during the Cretaceous: evidence ofGondwanan vicariance from biogeographical analysis. Proceedings of the RoyalSociety of London B 271, 2003–2009.

Tyson, R.V., Funnell, B.M., 1987. European Cretaceous shorelines, stage by stage.Palaeogeography, Palaeoclimatology, Palaeoecology 59, 69–91.

Upchurch, P., 2008. Gondwanan break-up: legacies of a lost world? Trends in Ecologyand Evolution 23, 229–236.

Upchurch, P., Hunn, C.A., Norman, D.B., 2002. An analysis of dinosaurian biogeography:evidence for the existence of vicariance and dispersal patterns caused by geologicalevents. Proceedings of the Royal Society of London B 269, 613–621.

Upchurch, P., Barrett, P.M., Dodson, P., 2004. Sauropoda, In: Weishampel, D.B., Dodson,P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition. University of California Press,Berkeley, Los Angeles and London, pp. 259–322.

Vartanyan, S.L., Garutt, V.E., Sher, A.V., 1993. Holocene dwarf mammoths fromWrangelIsland in the Siberian Arctic. Nature 362, 337–340.

Vartanyan, S.L., Arslanov, K.A., Tertychnaya, T.V., Chernov, S.B., 1995. Radiocarbondating evidence for mammoths on Wrangel Island, Arctic Ocean, until 2000 BC.Radiocarbon 37, 1–6.

Vartanyan, S.L., Arslanov, K.A., Karhu, J.A., Possnert, G., Sulerzhitsky, L.D., 2008. Collectionofradiocarbon dates on the mammoths (Mammuthus primigenius) and other genera ofWrangel Island, northeast Siberia, Russia. Quaternary Research 70, 51–59.

Venczel, M., Csiki, Z., 2003. New discoglossid frogs from the Latest Cretaceous of HaţegBasin (Romania). Acta Palaeontologica Polonica 48, 599–606.

Vianey-Liaud, M., 1979. Les Mammiféres montiens de Hainin (Paléocène moyen deBelgique). Part I: Multituberculés. Palaeovertebrata 9, 117–131.

Vianey-Liaud, M., 1986. Les Multituberculés thanetiens de France, et leur rapports avecles Multituberculés Nord-Americaines. Palaeontographica A191, 85–171.

Vickaryous, M.K., Maryanska, T., Weishampel, D.B., 2004. Ankylosauria, In: Weisham-pel, D.B., Dodson, P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition. University ofCalifornia Press, Berkeley, Los Angeles and London, pp. 363–392.

von Huene, F., 1932. Die fossil Reptil-Ordnung Saurischia, ihre Entwicklung undGeschichte. Monographen zur Geologie und Palaeontologie 4, 1–361.

Vremir, M.M., 2004. Fossil turtle finds in Romania — overview. Magyar Állami FöldtaniIntézet Évi Jelentése 2002, 143–152.

Vremir, M., Codrea, V., 2009. Late Cretaceous turtle diversity in Transylvanian and Haţegbasins (Romania). In: Bucur, I.I., Săsăran, E., Pop, D. (Eds.), Abstract Book. SeventhRomanian Symposium on Palaeontology. Presa Universitară Clujeană, Cluj-Napoca,pp. 122–124.

Wallace, A.R., 1860. On the zoological geography of the Malay Archipelago. ZoologicalJournal of the Linnaean Society 4, 172–184.

Weishampel, D.B., 1996. Fossils, phylogeny and discovery: a cladistic study of thehistory of tree topologies and ghost lineage durations. Journal of VertebratePaleontology 16, 191–197.

Weishampel, D.B., Jianu, C.-M., 1996. New theropod dinosaur material from the HaţegBasin (Late Cretaceous, Western Romania). Neues Jahrbuch für Geologie undPaläontologie Abhandlung 200, 387–404.

Weishampel, D.B., Jianu, C.-M., 1997. The importance of phylogeny in paleobiogeo-graphic analyses with examples from North American hadrosaurids and Europeantitanosaurids. Sargetia 17, 261–278.

Weishampel, D.B., Grigorescu, D., Norman, D.B., 1991. The dinosaurs of Transylvania:island biogeography in the Late Cretaceous. National Geographic Research andExploration 7, 68–87.

Weishampel, D.B., Norman, D.B., Grigorescu, D., 1993. Telmatosaurus transsylvanicusfrom the Late Cretaceous of Romania: the most basal hadrosaurid dinosaur.Palaeontology 36, 361–385.

Weishampel, D.B., Jianu, C.-M., Csiki, Z., Norman, D.B., 2003. Osteology and phylogeny ofZalmoxes (n.g.), an unusual euornithopod dinosaur from the latest Cretaceous ofRomania. Journal of Systematic Palaeontology 1, 65–123.

Weishampel, D.B., Barrett, P.M., Coria, R.A., Le Loeuff, J., Xu, X., Zhao, X.J., Sahni, A.,Gomani, E.M.P., Noto, C.R., 2004. Dinosaur distribution, In: Weishampel, D.B.,Dodson, P., Osmólska, H. (Eds.), The Dinosauria, 2nd edition. University ofCalifornia Press, Berkeley, Los Angeles and London, pp. 517–606.

Wiley, E.O., 1988. Vicariance biogeography. Annual Review of Ecology and Systematics19, 513–542.

Wiley, E.O., Siegel-Causey, D., Brooks, D.R., Funk, V.A., 1991. The compleat cladist.University of Kansas Museum of Natural History Special Publication 19, 1–158.

Worthy, T.H., Tennyson, A.J.D., Archer, M., Musser, A.M., Hand, S.J., Jones, C., Douglas, B.J.,McNamara, J.A., Beck, R.M.D., 2006. Miocene mammal reveals a Mesozoic ghostlineage on insular New Zealand, southwest Pacific. Proceedings of the NationalAcademy of Sciences of the USA 103, 19419–19423.

Xu, Z., Zhao, X.J., Lu, J.-C., Huang, W.-B., Li, Z.-Y., Dong, Z.-M., 2000. A new iguanodontianfrom Sangping Formation of Neixiang, Henan and its stratigraphical implications.Vertebrata PalAsiatica 38, 176–191.

You, H.L., Luo, Z.-X., Shubin, N.H., Witmer, L.M., Tang, Z.-L., Tang, F., 2003a. The earliest-known duck-billed dinosaur from deposits of late Early Cretaceous age innorthwest China and hadrosaur evolution. Cretaceous Research 24, 347–355.

You, H.L., Ji, Q., Li, J.L., Li, Y.X., 2003b. A new hadrosauroid dinosaur from the mid-Cretaceous of Liaoning, China. Acta Geologica Sinica (English Edition) 77, 148–154.

Ziegler, P.A., 1987. Evolution of the Arctic–North Atlantic and theWestern Tethys. AAPGMemoir 43 198 pp.